Best AI Certifications in 2021!

Those seeking a career in artificial intelligence (AI) can have an easy task when you have an AI certification from a reputed institute.

Artificial Intelligence (AI) is a skill that is programmed and designed so that the machines behave and think like a human. It is an imperative part of our daily lives and is used in many daily service areas. The introduction of artificial intelligence has brought the idea of ​​an error-free world and is gradually getting introduced in all fields to increase automation and provide accurate and faster results.

Becoming a certified AI professional is the key to a satisfactory career in the AI ​​domain and can be a tool to elevate your career for better projects and roles.

However, getting certified can be a huge task, as numerous institutes are offering AI certifications. So how do you choose the best amid the lot? Well! We have made that chore easy for you as we share the six best AI certifications in 2021, which will help you scale up your career.

• MIT AI Online Program: Massachusetts Institute of Technology(MIT) offers various online certification programs for artificial intelligence. The certification program by MIT offers practical knowledge in AI that will help you transform your organization into an efficient, sustainable, and innovative organization of the future.

This program will help you understand the design principles and applications of artificial intelligence in various industries. You will learn the various stages involved in AI-based product design and the basics of the machine and deep learning algorithms and apply these insights to solving practical problems. Your goal is to create an AI-based product proposal that can be submitted to your internal stakeholders or investors.

The program also offers the double perspective of management and AI, thus giving you a sound knowledge of AI-based technologies through the eyes of the business.

URL: https://bit.ly/3kPmXqH

• Machine Learning Engineering for Production (MLOps) Specialization by Coursera: Learn by the leaders. Created by none other than Andrew Ng, Machine Learning Program by Coursera has been cited as one of the best in the AI certification programs. Effective deployment of machine learning models requires more common capabilities in technical fields such as software engineering and DevOps. Machine learning engineering for production combines the basic concepts of machine learning with the functional expertise of modern software development and engineering roles.

Through this program, learners are introduced to basic machine learning ideas, including recognizing statistical patterns and data mining.

URL: https://bit.ly/3xZ9gt7

• AI Certifications by the United States Artificial Intelligence Institute (USAII™): The institute offers three major cross-platform (vendor-neutral) certification programs for aspiring AI professionals – for an engineer (students or working professionals with limited industry experience), consultant (for students having Master’s degree with limited industry experience and working professionals with more than two years of industry experience), and scientist (for working professionals with more than five years of industry experience) role in the field of AI. The need to evolve from being data-driven workflows to AI-driven has opened numerous job opportunities in the industry. Thus the three AI certifications, Certified AI Engineer (CAIE™), Certified AI Consultant (CAIC™), and Certified AI Scientist (CAIS™), provide the industry-relevant skills in AI.

All three programs are self-paced and offer preparatory study kits to the candidates comprising study books, videos, workshops, and practice code. Getting a vendor-neutral AI certification will have a multi-fold advantage. USAII™ claims to close the AI talent gap with these certifications across the globe.

URL: https://www.usaii.org

• IBM Applied AI Professional Certificate (Coursera): Designed by the global leader in Tech IBM, the Professional Certificate in Artificial Intelligence caters to professionals who want to work as AI developers. This plan will give you an in-depth understanding of AI technology, its applications, and use cases. You will be familiar with concepts and tools such as machine learning, data science, natural language processing, image classification, image processing, IBM Watson AI services, OpenCV, and APIs.

You will learn practical Python skills to design, build, and deploy AI applications on the web with this professional certificate without a programming background. These courses will also enable you to apply pre-built AI intelligence to your products and solutions.

URL: https://bit.ly/36RG4s7

• Coursera Artificial Intelligence Courses: Coursera provides a huge variety of certification programs and specializations in the field of AI. These programs have been designed in association with the world’s top universities and data science schools, including leaders in the AI industry. The programs created by Coursera introduce the students to the latest tools, concepts like Artificial Neural Networks, Deep Learning, TensorFlow, Python programming, and reinforcement learning.

The program ranges from beginner-level courses to senior-level programs for experienced AI professionals.

URL: https://bit.ly/2V2pHGn

• Enterprise AI and Machine Learning by Cloudera: The enterprise AI and Machine Learning program by Cloudera enables the learners to shift their focus from technology to the result, i.e., the outcomes, and thus empower the ongoing optimization in the field. Only through the industrialization of AI will you be able to shift focus from technology to outcomes, empower continuous optimization and learning across the enterprise, and turn data into predictions at any scale, anywhere.

With this program, an applicant can build and deploy and scale AI applications with a repeated industrialized methodology that can turn available data into insightful decisions.

URL: https://www.cloudera.com/about/machine-learning.html

These are some of the best AI certifications in the market. You can visit their website individually to gain more knowledge about each of the certifications and get started on your AI career path. You should carefully select the program that helps you best to learn essential AI skills and give you a forever valid digital badge (to brand you on social media like LinkedIn, Facebook, Instagram, and others). This will undoubtedly help you to find a job or promotion or increase your personal brand value as an AI expert.

Source Prolead brokers usa

No Code AI, No Kidding Aye!

When was the last time you did something meaningful for the first time? For me, that was in the last week of June’21. Just two weeks into my stint at Subex, I made an ML prediction model, my first one! Yes, I know that is nothing earth-shattering, but before you start rolling your eyes at my juvenile glee and start judging, let me tell you that I do not know how to code. I cannot code to save my life and the last time I wrote something that had a semblance of a code was two decades back when I was in graduate college.  So, yes, I am elated; not in the least because I made a very simple ML model, but because of the feeling of freedom and empowerment that I felt when I saw each one of my process steps in the pipeline lighting up in green and running its full course and culminating in successful output. In my mind, it is the kind of emotion one goes through when you have been handicapped for a lifetime, and then, one fine day, you get a bionic limb that liberates you and lets you walk freely once again.

The future potential of No Code/Low Code platforms

Gartner’s research points to the emerging trend that digital transformation initiatives have triggered an insatiable demand for custom software development. This, in turn, has ignited the emergence of citizen developers and citizen data scientists who are outside the traditional definitions of an IT developer or an ML developer. This paradigm shift has influenced the rise of no-code and low-code platforms. According to Gartner, on average, 41% of employees outside of IT – or ‘business technologists’ – customize or build data or technology solutions, and they are sticking their neck out confidently to say that that, by the end of 2025, 50% of all new low-code clients will come from business buyers that are outside the IT organization and 65% of all application development will be low code by 2024.

Enough of numbers for now. I guess we all get the drift – Low code/No code development platforms are the next big thing in software development. So, is this another technological development whose impact will stay limited to large, for-profit corporations and enterprises, or, will this have a greater, more purposeful bearing?

The profound impact of No Code platforms

In 2005, the Indian parliament passed a historic and landmark bill – The Right To Information Act, or RTI, as it is popularly known. This Act took the key that was needed to access data related to most of the day-to-day functioning of government bodies, from the hands of limited law enforcement and judiciary entities and passed it into the hands of the common man. Suddenly, everything changed. The fundamental societal framework that government functionaries could do whatever they wanted and get away with it unless someone with a lot of time, money, patience, courage, and determination could force the hands of the law using judicial process, was turned on its head. Anybody who was a citizen of India could pay a paltry sum and demand specific information from almost any government entity and was entitled to get that information. It was a true watershed moment for the democratic fabric of this great nation. It made governance more responsible and accountable, and it paved the way for numerous improvements and transparency at the grassroots level, and that, in the true sense, was a transformation. In case you are wondering how this is related to No Code AI platforms, hold on just a little bit more.

It was never the case that prior to the RTI Act getting passed, there was no data. Oceans of data existed, but what did not exist was a framework and structure in which that data could be leveraged by every citizen. Until then, data could only be sought and used by limited entities and institutions – legislature, law enforcement, and judiciary and often, they only sought and used data for purposes of utmost legal importance and priority. Compare that to the situation we have at hand in the private and business sector. Humungous amounts of data exist, but how can you leverage the true potential of it if the value extraction power is in the hands of a few limited people who need to have highly advanced, technical skills? Creativity and the power of imagination are not always ensconced inside technical or programming knowledge. There are so many business users who have fantastic ideas which never see the light of the day, either because it is not considered a priority, or simply because there’s just not enough bandwidth of expensive technical resources to spare to chase up every idea that is being tabled. In most cases, the process of innovation in organizations works like the highly competitive entrance examinations to prestigious colleges – it is a process of eliminating as many as possible, rather than retaining everyone who might have potential.

With RTI, anybody who wanted to check a hypothesis could ask for relevant data and had the right to receive that data within a stipulated timeframe. Suddenly, corruption became one step more arduous as the usage of data became truly democratized. NGOs and committed citizens who wanted to make real fundamental changes started gathering and foraging through data to identify patterns and anomalies and started asking questions, most of which were uncomfortable ones to answer but had the power to alter the basic fabric of the system.

AI has the power to change the way we eat, sleep, breathe and live. In business, it has the potential to transform fortunes and deliver exceptional customer experiences at scale, but the question we need to ask is – Given a choice, do we want to restrict and throttle this power, due to natural limitations, within the hands of a few people who need to have highly technical and specialized skills to know how to fully leverage it? Imagine the possibilities of this power in the hands of millions of creative and logical people who can dream up out-of-the-world applications for the data assets that we sit on without everybody having to be a software programmer or a technical data scientist. That would be true democratization of data and AI, and No-Code platforms are paving the way for this revolution.

Note: This blog post is part one of a two-part series on the No-Code AI platform revolution. The next part will cover the challenges of the AI model building that No-Code AI platforms address.

Till then, stay tuned…

Source Prolead brokers usa

DSC Weekly Digest 3 August 2021

There were several interesting announcements this week about the Rise of the Metaverse. Depending upon who you talk to, it’s the next big thing, Internet 3.0, Snow Crash careening into Ready Steady Go, with vibes of Tron thrown in for good measure. It’s Virtual Reality 2.0. And it’s coming to a screen near you tomorrow … or maybe in fifty years. You can never tell with these kinds of things.

There is a certain innate similarity between virtual reality and self-driving cars. To hear the press releases from either 1999 or 2015, VR and autonomous vehicles were literally just around the corner, an engineering problem, not a conceptual problem. By 2021, the first truly consumer autonomous vehicles were supposed to be coming off the assembly lines, and VR should have been achieved by now. Instead, AVs are still at least a decade away and truly immersive, fully interactive VR should be a thing.

Now, anyone who games regularly can tell you that immersive realities are definitely here – so long as you’re very careful to constrain how far out of the box someone can go. Anyone who’s played Halo or Overwatch or even Dead Red Redemption can tell you that the games are becoming quite realistic, and arguably games such as the Sims (version x) attest to the ability to have multiple individuals within a given simulation.

As with AVs, the challenge ultimately isn’t engineering – it’s social. Second Life explored the themes of virtual reality in a social sense. What happened afterward was simple: people discovered that virtual conversations and virtual sex with virtual avatars was, at the end of the day, boring and more than a little creepy. It was like going to a bar without any alcohol. 

We enjoy games precisely because we are, to quote Terry Pratchett, narrative creatures. We are natural storytellers, and we love both being told and participating in stories. We love pitting ourselves against others, seeing ourselves as fighting the good fight or solving deep mysteries that would have stymied Sherlock Holmes. Psychologists also talk about the dangers of escapism, but games are attractive primarily because most IRL stories are not very exciting.

There’s some significant money to be made in Extended Reality (XR), but it’s important to understand there is that for it to truly work, XR needs to concentrate as much on the metadata, the story, as it does on the various communication protocols and representations.

The latter is not insignificant, mind you. The virtual world is the quantum cognate of the real world. Identity and uniqueness are intrinsic to the physical world, and creating duplicates that travel through real space and time is a nearly insurmountable problem. In the virtual world, however, uniqueness and identity are simply abstract concepts, and creating copies that can persist for any significant length of time can prove difficult at best (this is what blockchain is supposed to do, but we’re discovering the very real energy costs in even approximating uniqueness).

Yet, ultimately, the real challenge will come when the various players in this space recognize that without compelling content where immersion means that people become a part of the narrative, not simply an avatar walking around stiffly in a pretty landscape, XR will fail. I’d also like to believe that ultimately it will take agreement on standards for all of the fiddling bits, like identity management, concurrency, data flows,, and so forth, to all come together so that moving from one narrative to another becomes feasible (or even makes sense), but I suspect that will only come once the landscape has become nearly irrevocably fractured. There are too many people with dollar signs in their eyes at this stage to expect any difference.

What does this mean to data science? Easy – a game is simply a simulation with a better plot. AI is intimately tied to the concept of Metaverse, and will only become more so over time.

Some Recent Changes

There are a couple of new changes to the newsletter. The first is changing the DSC article listings so that they show authors and publication dates. We’re proud of our writers, and I feel that by posting who wrote what will make it easier for you as a writer to go to those writers that you enjoy most, as well as helping you discover new and different viewpoints. Clicking on the writer links will give you a feed showing all of their previous articles.

Another, more subtle change is that as a member of DSC clicking on a Tech Target article will take you to the article without triggering the paywall. You can now enjoy more of our parent company’s content, and get perspective from industry leaders. It will also help us track what’s most important to you, our readers. Note that you can only see TechTarget content when coming from the Newsletter or from the DSC site itself.

In media res,

Kurt Cagle
Community Editor,
Data Science Central

To subscribe to the DSC Newsletter, go to Data Science Central and become a member today. It’s free! 

Source Prolead brokers usa

As the Digital Workforce Changes, So Will the Economy

Image credit: Unsplash. 

This is an Op-Ed about the relationship between remote work, the economy and our physical communities health and well-being. 

As people are glorifying the remote work WFM trend, for some reason the dark side of digital transformation isn’t being told. Yet technology, data science and the future of artificial intelligence impacts how society functions, how economies evolve and now even how neighborhoods and communities will ‘feel’. 

There is a very stark reality that’s not being talked about during the pandemic very much. It’s actually how remote work (the WFM movement) is bad for the economy. While it gives workers more freedom and saves on time and productivity, the powerful side effect is less spending for local businesses and small businesses that depend upon commuters.

While people like to blame the pandemic for this, it’s actually digital transformation leaders that permanently upend the small and medium-sized business (SMB) sector here as accelerating the WFM remote work trend. While companies like Zoom, Microsoft and others thrived during the transition, many SMBs and local business are already gone including restaurants, coffee shops, independent retailers, dry cleaners, etc… you get the idea.

Basically all the things commuters do on their daily runs to and from the office could be somewhat upended. So is the real story that remote work is bad for the economy or that digital transformation has a dark side on the nature of the economy?

The Impact of Remote Work on the Economy is Underreported 

In 2020, the number of people working from home nearly doubled, to 42% of America’s workforce, according to the Bureau of Labor Statistics. Several companies in the tech sector are giving employees the option to consider hybrid or full on forever remote work. Less commuting also means lower consumption of oil, among many other things. So the impact of remote work on the economy will in some ways have better productivity for Big Tech giants but severely impact the sustainability of small businesses who already have tight margins.

CNN goes on to state that you pay train conductors’ salaries with your subway fare. The dry cleaner by the office and the coffee shop around the corner all count on workers who have been largely absent for nearly a year and a half. The pivot to hybrid and remote work is not very well understood, since it’s so recent and new. But the dark sides of digital transformation are certainly being suppressed from the mainstream news.

Those train tickets or lattes really do add up and taking them away can also hurt local transportation networks and the future of urban commuting itself. The WFM movement could accelerate the profits of certain companies while disrupting others. For many SMBs this is really sad since the pandemic marks a permanent shift of how their business model will no longer be viable, this impacts local economies, the destruction of neighborhood commerce and SMBs.

Is the Digital Workforce Harmful to Local Economies?

Companies like Microsoft, Zoom, or Peloton will argue the pandemic is to blame; they were just opportunists who helped us through the transition. But digital transformation will, in this way, accelerate wealth inequality and low paid service jobs will be impacted by those lost small businesses. The huge corporations profiting during the lockdowns and pandemic work situations hugely accelerated their revenue and seek to keep the switch of all things digital and remote work as a permanent trend.

In the debate of whether remote work harms or helps the economy, a lot is lost in translation.

To put this in perspective, New York’s public transport system is the largest in the nation and at the heart of the city’s economic power. Before Covid, it brought in nearly $17 billion in revenue. But with ridership still depressed, revenue predictions have been slashed too. The Metropolitan Transit Authority received nearly $4 billion in government funding through the CARES Act, but fare and toll revenues aren’t expected to come back to their previous levels until 2023, according to a report from the Office of the New York State Comptroller earlier this year (CNN).

The rapid spread of the Delta variant and the coronavirus becoming endemic basically means the digital transformation natives are validated, they will further disrupt the local SMB sector all over the U.S. and to some extent, most urban centers in the world. This is a huge transfer of wealth and for local economies will have real and significant costs that aren’t being monitored or regulated.

Corporate Individualism and the Push for the Digital Workforce Could Upset Communities and Small Businesses 

What’s for the individual worker (actually the Big Tech corporation) is not always good for the collective. It’s great to have the freedom as a knowledge worker to work remotely, we don’t want to hurt local businesses or impact the economy with our decisions. But just remember that when digital transformation occurs, there’s a dark side of the corresponding disruption of the way things were.

The pandemic has been a grand global experiment in the costs and benefits of a remote workforce. But long before the coronavirus hit, many people worked from outside offices. When opportunistic digital value providers can benefit from the pandemic as we have seen with software services, advertising, social media, gaming or entertainment – what happens in the real world isn’t talked about as much.

The artificial intelligence, cloud computing and software as a service models at the backbone of these transformative new services that cater to the remote worker, and the more remote work wins in the equation of the digital workforce, the harder it becomes for many brick-and-mortar retailers, local shops, services and businesses to even have the opportunity to make a living. In this way the best companies in AI, cloud and subscription software services are disrupting the rest of society and taking their money (wealth distribution, wealth inequality rising and exploiting their advantage against less digitally native companies).

What if Digital Transformation is Not All Good?

We used to fear what Amazon was doing to local retailers, but now it’s Microsoft as well. It’s Zoom, Peloton and so many others who have triumphed when many of us are at our most vulnerable. Hybrid work, remote work and freelancing is likely here to stay to some extent in a post-pandemic world.

More than two-thirds of professionals were working remotely during the peak of the pandemic, according to a new report by work marketplace Upwork, and over the next five years, 20% to 25% of professionals will likely be working remotely. 

By late April, more than half of all workers, accounting for more than two-thirds of all U.S. economic activity, said they were working from home full-time. According to Nicholas Bloom, an economist at Stanford University who has studied remote work, only 26 percent of the U.S. labor force continues to work from their job’s premises.

Some, including several Silicon Valley giants, have announced that they will allow employees to work from home permanently. Yet huge swathes of the labor force are unable to work remotely, and experts say these developments could have profound implications for the economy, inequality, and the future of big cities. Of course Silicon Valley love the idea of making remote work as permanent as possible, they are the main beneficiaries of the new normal!

So what will the new normal look and feel like? The destruction of local communities? “Job ads increasingly offer remote work and surveys indicate that both workers and employers expect work from home to remain much more common than before the pandemic,” Goldman Sachs economists said in a note to clients.

Digital transformation leads have accelerated the WFM reality with their products, services and getting us hooked to new ways of life that involve spending less on local businesses and going to the office much less, if at all. Now employees crave more hybrid work convenience or even a full transition to remote life. The danger to local small businesses and neighborhood commerce is being totally ignored.

Indie Retailers and Local Businesses will be Severely Impacted by Remote Work

In Capitalism, we just assume a natural selection of how businesses compete for advantages is fair or par for the course. But remote work stacks the cards even further against the little guys, the indies, the small Mom and Pops store, the family business, the struggling immigrants, the low paid workers of retail and so forth. Remote work will decimate the local economy and significantly degrade the experience and immersion of neighborhoods and communities, and in 2021 we aren’t even talking about it.

If lockdowns weren’t positive for local businesses who had no customers, what will the WFM be for them honestly speaking on one third to half less foot traffic? Small businesses that used to depend on traffic from offices, won’t just struggle, most of them will gradually go out of business. 

Remote work is mainly the luxury of knowledge workers, but here we are widening the gap of white and blue collar to an absurd degree. It basically guarantees that much of the lower middle class falls lower on the spectrum while some in the upper middle class get a lot of more benefits. Remote work weirdly augments inequality and if I was an economist, I’d be ringing the alarm bells. The WFM movement is great, until you realize not everyone will be along for the ride.

The remote work trend is likely to accelerate wealth inequality and lower the quality of life and well-being in many communities as the demand for many kinds of SMBs dries up due to less foot traffic in the new normal of the digital workforce. 

It’s perhaps the weirdest untold story of the pandemic people ignore unless it impacts them personally. The WFM trend is not all good, and its economic impacts to the SMB sector is going to be significant and likely, permanent. 

Source Prolead brokers usa

Do we need AutoML… or AutoDM (Automated Data Management)?

Instead of focusing on “Automated Machine Learning” or AutoML, maybe we should focus on “Automated Data Management” or AutoDM?

You probably know that feeling. You start a blog with some ideas to share, but everything changes once you get started. That’s what happened with this blog.  I discussed the promise and potential of Automated Machine Learning (AutoML) in my blog “What Movies Can Teach Us About Prospering in an AI World“.  It seems quite impressive.

So, I decided to conduct a LinkedIn poll to garner insights from real-world practitioners, folks who can see through the hype and BS (and that’s not Bill Schmarzo…or maybe it should be) about the potential ramifications of AutoML.  It was those conversations that lead to my epiphany. But before I dive into my epiphany, let’s provide more on AutoML.

“Automated machine learning (AutoML) automates applying machine learning to real-world problems. AutoML covers the complete pipeline from the raw data to the deployable machine learning model. The high degree of automation in AutoML allows non-experts to create ML models without being experts in machine learning. AutoML offers the advantages of producing simpler solutions, faster creation of those solutions, and models that often outperform hand-designed models.[1]” See Figure 1.

Figure 1: Image sourced from: “A Review of Azure Automated Machine Learning (AutoML)”

Man, that is quite a promise. But here’s the AutoML gotcha: to make AutoML work, data experts need to perform significant data management work before getting into the algorithm selection and hyperparameter optimization benefits of AutoML.  This includes:

• Data Pre-processing which includes data cleansing (detecting and correcting corrupt or inaccurate records), data editing (detecting and handling errors in the data), and data reduction (elimination of redundant data elements).
• Data wrangling which transforms and maps data from one “raw” data format into a format that is usable by the AI/ML models.
• Feature Engineering which is the process of leveraging domain knowledge to identify and extract features (characteristics, properties, attributes) from raw data that are applicable to the problem being addressed.
• Feature Extraction involves reducing the number of features or variables required to describe a large set of data. This likely requires domain knowledge to identify those features most relevant to the problem being addressed.
• Feature Selection is the process of selecting a subset of relevant features (data variables) for use in model construction. Again, this likely requires domain knowledge to identify those features most relevant to the problem being addressed.

That’s a lot of work to do before even getting into the AutoML space.  But us old data dogs already knew that 80% of the analytics work was in data preparation.  It’s just that today’s AI/ML generation needs to hear that, and who better to deliver that message than one of the industry’s AI/ML spiritual leaders – Andrew Ng.

Here is a must watch video from by Andrew titled “Big Data to Good Data: Andrew Ng Urges ML Community to Be More Data…”.  There are lots of great insights in the video, but what struck me was Andrew’s own epiphany on the critical importance of spending less time tweaking the AI/ML models (algorithms) and investing more time on improving the data quality and completeness that feeds the AI/ML models. Andrew’s message is quite clear:  while tweaking the AI/ML algorithms will help, bigger improvements in overall AI/ML model performance and accuracy can be achieved by quality and completeness improvements in the data that feed the AI/ML algorithms (see Figure 2).

Figure 2: Transitioning from Algorithm-centric to Data-centric AI/ML Model Development

And note that those improvements in data quality and completeness that feeds the AI/ML models will benefit all AI/ML models that use that same data!  Sounds a lot like the Schmarzo Economic Digital Asset Valuation Theorem – the economic theorem on sharing, reusing, and refining of the organization’s data and analytic assets.

In the video, Andrew shared hard data with respect to improvement in results from tweaking the model (algorithm) versus improving data quality and completeness (see Figure 3).

Figure 3: Improving the Code versus Improving the Data

In the three use cases in Figure 4, there was literally no improvement in AI/ML model accuracy and effectiveness from tweaking the AI/ML models.  However, efforts applied against improving the data yielded quantifiable improvements, and in one case, very significant improvements!

Figure 4 shows the LinkedIn poll results where I asked participants to select the option they felt was most true about AutoML (sorry, only 4 options are available on LinkedIn).

Figure 4: LinkedIn AutoML Poll

If we factor the “All of the Above” choice with the top two choices, we get the following results:

• 62% of respondents feel AutoML will help automate data science model development
• 56% of respondents feel AutoML will enable business users to build their own ML models

Unfortunately, not having a “None of the Above” option was unfair because the results of the poll differ from poll comments. Here is my summary of those comments:

• AutoML will not be replacing data scientists anytime soon. However, AutoML can help jumpstart the Data Science process in ML model exploration, model selection, and hyperparameter tuning.
• AutoML will not suddenly turn business analysts into data scientists. That’s because ~80% of the ML model development effort is still focused on data preparation. To quote one person, “AutoML by untrained users would be like giving an elite athlete training plan and diet to average people and expecting elite results.”
• AutoML will be even more lacking as Data Scientist’s data preparation work evolves to semi-structured (log files) and unstructured data (text, images, audio, video, smell, waves).
• Realizing the AutoML promise will require a strong metadata strategy and plan.
• AutoML could help in AI/ML product management as the number of production ML models grows into the hundreds and thousands. But AutoML would need an automated set-up to monitor and correct for ML data drift while in production.
• Automating the ML process is just a small step. AutoML results need to be explainable to help in the evaluation of the analytic results using techniques such as SHAP or CDD.
• AutoML is a commodification of the loops and utilities that ML folks run through various ML algorithms, tune hyper-parameters, create features, and calculate metrics of all kinds.
• AutoML can be a great tool to get align teams around an organization’s ML aspirations. A field only flourishes when everyone from every discipline can use it to try different ideas.
• For AutoML to be successful, it is critical important to scope, validate, and plan the operationalization of the problem that one is trying to solve (e.g., Is the target variable here *really* what you want to model? Are all of the inputs available in a production environment? What decisions will this model support? How will you monitor the ongoing accuracy and usage of the model? How will you govern changes to the model, including commissioning and decommissioning it?). Hint, see Hypothesis Development Canvas?
• Finally, is AutoML a marketing ploy by cloud vendors to broaden their appeal to include enabling business users to build their own ML models?

I suggest that you check out the chat stream.  The comments were very enlightening.

My takeaway is that the concept of AutoML is good, but scope of the AutoML vision is missing 80% of the AI/ML model development and operationalization – providing high quality and complete data that feeds the AI/ML models. Figure 5 from “Big Data to Good Data: Andrew Ng Urges ML Community to Be More Data…” nicely summarizes the broader AutoML challenge with respect to data management.

Figure 5: Scope of What AutoML Needs to Address

Instead of focusing on “Automated Machine Learning” or AutoML, maybe we should focus on “Automated Data Management” or AutoDM?

Now that’s a thought…

[1] Wikipedia, AutoML https://en.wikipedia.org/wiki/Automated_machine_learning

Source Prolead brokers usa

Eight Tips to Manage Your Remote Team in 2021

The concept of remote work was alien to almost every professional until March 2020 when the World Health Organization declared COVID-19 a deadly pandemic.

After the severity of the situation increased, every organization, every professional had to adapt to remote working mandatorily as the governments across the world announced complete lockdowns and major restrictions on the movement.

With no prior experience and understanding of how remote work is carried out at scale, organizations, business leaders, HR leaders, senior and mid-level managers, as well as executive-level staff struggled to find the right spot between communication, executive, and innovation.

The first half of 2020 was a learning curve for everyone while they accepted and understood the work from home reality, and adapted accordingly. But 2021 is definitely the year where new norms of working are getting defined and remote working is going to play a really big part. According to a survey done by Tecla, 85% of managers believe that teams with remote work will become the future of work.

“The future we envision for work allows for infinite virtual workspaces that will unlock social and economic opportunities for people regardless of barriers like physical location. It will take time to get there, and we continue to build toward this.” – Andrew Bosworth, VP Facebook Reality Labs

Since remote work is the future, every manager should know the nuances, best practices, and strategies of managing a remote team which is what we are going to discuss in this article.

[Tried & Tested] Tips & Strategies to Manage a Remote Team in 2021

All the strategies and tips listed below to manage a remote team are based on the biggest struggles of working remotely.

[Source: Buffer]

Communicate All the Kinks Out

It’s easier to communicate but it’s the hardest to communicate clearly and effectively. The better the communication, the easier it will be for you to set expectations, communicate deadlines and project specifics.

Use deliberate and structured communication whenever communicating with your team members. If possible, have weekly work meetings on Mondays to set goals for the week and then a sort of debriefing and fun meetings on Fridays for catching up, understanding progress on projects, weekend plans, etc. This way, you hit the sweet spot between formal, productive, and informal communication.

Also, the key to successfully conducting these meetings is video calls, so always make sure your team is switching their videos on and proactively participating in conversations. To be more effective, invest time in your team members – have a habit of conducting 1-1 meetings with them on a regular basis to understand how they are doing.

“Technology now allows people to connect anytime, anywhere, to anyone in the world, from almost any device. This is dramatically changing the way people work, facilitating 24/7 collaboration with colleagues who are dispersed across time zones, countries, and continents. ” — Michael Dell, Dell

Establish Complete Feedback Loops

Having feedback loops is quite important for remote teams since everyone is working remotely and in different time zones.

Establishing a feedback culture will help you provide support to your team members at an individual level, identify pain areas in operations, stay ahead of any potential conflict, and build meaningful relationships. All this will help improve your team’s overall performance.

Tips on How to Build Feedback Culture

• Make it a part of your process from day 1
• Create a safe environment for your team to express their feedback and concerns openly
• Train your team to give receive feedback – it is essentially a skill
• Use different feedback channels like 1-on-1 meetings, 360 feedback, anonymous feedback, etc.

Boundaries are Productive

After working from home for over a year, all of us have realized that the lines between work life and personal life can easily get blurred when you are working from home. So, it’s important to set some healthy boundaries for all your remote team members to avoid extra, unnecessary stress, and burnout.

For example, recently, Bumble – a dating app company, announced a week-long holiday for all their employees to avoid burnout.

A Few Ways to Go About Setting Boundaries

• Limit availability
• Ask them to avoid connecting their professional accounts on their personal devices
• Encourage wellness and self-care activities like mindful meditation breaks
• Share about personal interests and hobbies or any other non-work talks to keep it light
• Most importantly, don’t schedule too many meetings

Invest in Right Tools & Technologies

Since all of your team members are scattered across different cities, countries, and even continents, it becomes imperative to invest in the right tools and technologies that enable effective timely collaboration.

Things to Keep in Mind While Choosing Tools for Your Team

• Consider all the use cases and then hunt for the right product.
  

Choose future-proof tools that enable digital transformation for your organization. For example, if you were planning to invest in SaaS tools that enhance customer experience, then instead of going for software that enables better communication between your team and customers, invest in software that adds to their experience directly like a product tour software for customers’ onboarding. 

Here are some product tour examples and a guide on how to make the most out of such software.

• Review your process and needs and then choose accordingly
• Do your research, thoroughly
• Get the tools customized, if needed
• Pick the tools that allow you to create an integrated ecosystem
• Train your team

“The whole conversation is about how remote work is different, instead of being about the amazing tools we have at our disposal that remote teams and non-remote teams are able to use at any time. We have this opportunity to have a lot more freedom in our environment compared to when we had to be in an office, or even in school, 40 hours per week.” — Hiten Shah, FYI

Mandatory Monthly or Quarterly Holidays

If your team members haven’t opted for any holidays in the last few months, then make sure they take holidays, mandatorily. Just because we all are working from home and can’t travel, for the time being, doesn’t mean we don’t need holidays.

So make sure to keep an eye out for your team by encouraging them to time off even if they think they don’t need it.

Lastly, Acknowledge & Celebrate Milestones and Hard Work

Celebrating achievements and milestones is quite necessary to keep everyone motivated. Before remote working became the norm, it used to be easy to gather everyone and celebrate individual achievements, company-wide milestones. But when everyone is working from their own spaces, celebrating achievements gets forgotten easily. 

So, make sure to make it a habit to celebrate and acknowledge your team members whenever they are due.

“Now that companies have built the framework – and experienced the cost and time savings associated with it – there’s no real reason to turn back.”  –  Mark Lobosco

Source Prolead brokers usa

Charticulator : Creating Interactive Charts Without Code

Background

Many areas of AI continue to show innovation at an exciting pace.

For example, today, generative methods are on my radar. So, its nice to see innovation in an area which is stable and mature.

Charticulator from Microsoft research is a free and open source tool for creating bespoke Interactive Chart designs without the need for coding

A gallery of charticulator charts gives you an example of what’s possible

What is the innovation?

While its visually interesting, its important to understand what is the innovation here

Charticulator is an interactive authoring tool that allows you to make custom and reusable chart layouts.

Typically, chart creation interfaces force authors to select from predetermined chart layouts, preventing the creation of new charts.

Charticulator, on the other hand, converts a chart specification into mathematical layout constraints and uses a constraint-solving algorithm to automatically compute a set of layout attributes to actualize the chart.

It enables the articulation of complex marks or glyphs and the structure between these glyphs, without the need for any coding or constraint satisfaction knowledge.

The capacity to build a highly customized visual representation of data, tuned to the specifics of the insights to be given, increases the chances that these insights will be recognized, understood and remembered in the final implementation.

This expressiveness also provides a competitive advantage to the creator of this visual representation in a landscape saturated with traditional charts and graphs.

The charticulator approach lies at the intersection of three ideas:

• People make charts by hand drawing or programming, in addition to using interactive charting tools. Because they cannot tie many aspects of data to graphical elements, illustration tools are insufficient for building custom charts.
• On the other hand, most interactive charting tools based on templates require chart authors to first select from a set of basic chart types or templates, such as bar, line, or pie charts and offer limited customization possibilities beyond that.
• Meanwhile, creating a powerful custom chart with a library like D3.js or a declarative language like Vega gives you a lot of control over how data is encoded to graphical markings and how they are laid out. However, this method is only available to a restricted set of people with advanced programming skills.

So, you could perhaps think that the innovation of charticulator lies in democratizing the custom chart approach and making it easier with no code / low code – and therein lies the innovation IMHO

From a research standpoint, the following are the work’s significant research contributions:

• Charticulator’s design framework, which may be used to create a variety of reusable chart layouts.
• The implementation of Charticulator, a design framework that achieves the design framework by transforming the chart specification into layout constraints and incorporates a constraint-based layout algorithm and a user interface that allows interactive chart layout definition.
• The results of three types of testing: a gallery of charts to demonstrate Charticulator’s expressiveness, a chart reproduction study, and a click-count comparison versus three current tools

I think this is an exciting innovation – and also great that it is free and open source

Source Prolead brokers usa

Big Data: Effective tips to success

                                                              Image source: Data Science Blog

Can data especially big data be considered as the new gold? Considering the pace at which data is evolving all across the globe, there is little question. Big data contains huge information and we can extract them by performing big data analysis. Consider the following: 

• Netflix saves $1 billion per year on customer retention only by utilizing big data.
• Being the highest shareholder of the search engines market, Google faces 1.2 trillion searches every year, with more than 40,000 search queries every second!
• Additionally, among all the google searches. 15% of those are new and are never typed before, leading to the fact that a new set of data is generated by Google continuously regularly. The main agenda is to convert data into information and then convert that information into insights. 

Why need a Proper Big Data Analysis Strategy?

Organizations were storing tons of their data into their databases without knowing what to do with that data until big data analysis became a completely developed idea. Poor data quality can cost businesses from $9.7 billion to 14.2 millions every year. Moreover, poor data quality can surely lead to wrong business strategies or poor decision-making. This also results in low productivity and sabotages the relationship between customers and the organization, causing the organization to lose its reputation in the market.  

To deter this problem, here is a list of five things an enterprise must acquire in order to turn their big data into a big success:

Strong Leadership Driving Big Data Analysis Initiatives  

The most important factor for nurturing data-driven decision-making culture is proper leadership. Organizations must have well-defined leadership roles for big data analytics to boost the successful implementation of big data initiatives. Necessary stewardship is crucial for organizations for making big data analytics an integral part of regular business operations. 

Leadership-driven big data initiatives assist organizations in making their big data commercially viable. Unfortunately, only 34% of the organizations have appointed a chief data officer to handle the implementation of big data initiatives. A pioneer in the utilization of big data in the United States’s banking industry, Bank of America, specified a Chief Data Officer (CDO) who is responsible for all the data management standards and policies, simplification of It tools and infrastructures that are required for the implementation, and setting up the big data platform of the bank. 

Invest in Appropriate Skills Before Technology

Having the right skills are crucial even before the technology has been implemented: 

• Utilize disparate open-source software for the integration and analysis of both structured and unstructured data. 
• Framing and asking appropriate business questions with a crystal-clean line of sight such as how the insights will be utilized, and 
• Bringing the appropriate statistical tools to bear on data for performing predictive analytics and generating forward-looking insights. 

All of the above-mentioned skills can be proactively developed for both hiring and training. It is essential to search for those senior leaders within the organization who not only believe in the power of big data but are also willing to take risks and perform experimentation. Such leaders play a vital role in driving swift acquisitions and the success of data applications. 

Perform Experimentation With Big Data Pilots

Start with the identification of the most critical problems of the business and how big data serves as the solution to that problem. After the identification of the problem, bring numerous aspects of big data into the laboratory where these pilots can be run before making any major investment in the technology.  Such pilot programs provide an enormous collection of big data tools and expertise that prove value effectively for the organization without making any hefty investments in IT costs or talent. By working with such pilots, implementation of these efforts at a grassroots level can be done with minimal investments in the technology. 

Search For a Needle in an Unstructured Hay 

The thing that always remains on the top of the mind of businesses is unstructured and semistructured data – the information contained in documents, spreadsheets, and similar non-traditional data sources. According to Gartner, data of organizations will evolve by 800% in the upcoming five years and 80% of that data will be unstructured. There are three crucial principles associated with unstructured data. 

• Having the appropriate technology is essential for storing and analyzing unstructured data. 
• Prioritizing such unstructured data that is rich in information value and sentiments. 
• Extracting relevant signals must be done from the insights and must be combined with structured data for boosting business predictions and insights.

Incorporate Operational Analytics Engines

 One potential advantage that can be attained by using big data is the capability of tailoring experiences to customers based on their most up-to-the-minute behavior. Businesses can no longer extract the data of last month, analyze that data offline for two months, and act upon the analysis three months later for making big data a competitive benefit.

Take, as an example, loyal customers who enter promotional codes at the time of checkout but discover that their discount is not applied to result in a poor customer experience.

Businesses need to shift their mindset of traditional offline analytics to tech-powered analytic engines that empower businesses with real-time and near-time decision-making, acquiring a measured test and learn approach. This can be achieved by making 20% of the organization’s decisions with tech-powered analytical engines and then gradually increasing the percentage of decisions processed in this way over time as comfort grows about the process. 

Final Thoughts 

In this tech-oriented world and digitally powered economy, big data analytics plays a vital role in the proper navigation of the market and to come up with appropriate predictions as well as decisions. Organizations must never ignore understanding patterns and deterring flows. especially as enterprises deal with different types of data each day, in different sizes, shapes, and forms. The market of big data analytics is growing dramatically and will reach up to $62.10 billion by the year 2025. Considering that progression, 97.2% of the organizations are already investing in artificial intelligence as well as big data. Hence organizations must acquire appropriate measures and keep in mind all the crucial above-mentioned tips for turning their big data into big success to stay competitive in this ever-changing world.

Source..

Source Prolead brokers usa

Key Attributes in ER Diagrams

• Overview of different types of keys used in E-R diagrams.
• How to establish a primary key from a set of alternatives.
• Composite, superkey, candidate, primary and alternate keys explained.
• Other keys you may come across include foreign and partial keys.

If you’re unfamiliar with entities and attributes, you may want to read Intro to the E-R Diagram first.

The ER diagram is a way to model a database in an organized and efficient way. A key is a way to categorize attributes in an E-R diagram. When you first start making E-R diagrams, the number of different choices for keys can be overwhelming. However, the goal of the E-R diagram is to create a simplified “bird’s eye” view of your data. A judicious choice of keys helps to achieve that goal. Although there are many different types of keys to choose from in a set of data, relatively few will actually make it to your finished diagram.

Composite Keys

In general, keys can be single-attribute (unary) or multi-attribute (n-ary). A composite key requires more than one attribute. If a key is composite, like {state,driver license#}, a composite attribute can be made from the separate parts. In the early stages of database design, like E-R diagram creation, entity attributes are often composite or multi-valued. However, these may create problems down the road and need to be handled in specific ways when you translate into an actual database [1]. 

Superkey, Candidate, Primary and Alternate Key

 A superkey of an entity set is an attribute, or set of attributes, with values that uniquely identify each entity in the set. For example, a DMV database might contain the following information:

In this example, {License #} is a superkey as it is a unique identifier. {Make,Model,Owner,State,License#,VIN#} and {State,License#,VIN#} are also superkeys. On the other hand, {Owner} or {Make,Model,Owner} are not superkeys as these could refer to more than one person [2].

A candidate key is a minimal super key that uniquely identifies an entity. Minimal superkeys have no unnecessary attributes; In other words, superkeys that don’t have subsets that are also superkeys.  For example, {State,License#} or {VIN} in the above set are possible choices for candidate keys.  

One you have identified all the candidate keys, choose a primary key. Each strong entity in an E-R diagram has a primary key. You may have several candidate keys to choose from. In general, choose a simple key over a composite one. In addition, make sure that the primary key has the following properties [3]:

1. A non-null value for each instance of the entity.
2. A unique value for each instance of an entity.
3. A non-changing value for the life of each entity instance.

In this example, the best choice to identify a particular car is {VIN}, as it would never change for the lifetime of the vehicle. The first digit of a driver license number will change when a name change occurs, so this does meet the requirements of property 3 above. In addition, {VIN} is the logical choice because it is directly associated with the car. If an ownership change would occur, the VIN would stay the same. An alternate key is any candidate key not chosen as the primary key. For this example, {State,License#} is an alternate key. A partial key identifies a weak entity. Weak entities–those that rely on other entities–do not have primary keys [4]. Instead, they have a partial key–one or more attributes that uniquely identify it via an owner entity. 

When the word “key” is used in an E-R diagram, it usually refers to the primary key for an entity [5]. Show the primary key by underlining the attribute. 

Dashed underlines indicate partial keys. 

Other Keys You May Come Across

Foreign keys are not used in E-R models, but they are used in relational databases to indicate an attribute that is the primary key of another table. Foreign keys are used to establish a relationship when both tables have the same attribute [5].

A secondary key is used strictly for retrieval purposes and accessing records [5]. These keys do not have to be unique and are typically not included in an E-R diagram. The term secondary key is also occasionally used as a synonym for alternate key [6].

References

[1] Entity-Relationship modeling

[2] Relational Model and the Entity-Relationship (ER) Model

[3] Primary and Foreign Keys

[4] Entity-Relationship Diagram Symbols and Notation

[5] The Entity-Relationship Model

[6] Database Design ER

Source Prolead brokers usa

How Data Science And Machine Learning Works To Counter Cyber Attacks

We are all aware of the heinous cyber-attack that took down more than 200,000 systems in 150 countries in only a few days in May 2017. This was found by the National Security Agency (NSA) and was nicknamed “WannaCry,” which exploited a vulnerability and stole important resources before being distributed online. 

After successfully accessing the computer, it encrypted the machine’s contents and rendered them unreadable. Now, victims of the assault were informed they needed to acquire special decryption software to retrieve their stolen material. Furthermore, the attackers marketed this software.

This ransomware outbreak targeted both people and big organizations, including the United Kingdom’s National Health Business, Russian banks, Chinese schools, the Spanish telecommunications company Telefonica, and the US-based transportation service FedEx. 

The overall losses were estimated at $4 billion. Other forms of cyber intrusions, such as crypto jacking, which are more subtle and less destructive but costly, are on the rise. Even high-profile firms with sophisticated cybersecurity processes are vulnerable. 

A recent panic at Tesla in 2018 was averted owing to a diligent third-party team of cybersecurity specialists. As a result, there were over 11 billion malware infections in 2018. That is a major problem that cannot be solved solely by humans.

Fortunately, this is where machine learning may come in handy.

How Machine Learning Helps to Boost Cybersecurity? 

Machine learning is a subset of artificial intelligence that makes assumptions about a computer’s behavior by using algorithms from prior datasets and statistical analysis. It allows the computer to modify its operations and even execute functions for which it was not expressly intended. Thus, the role of ML and AI in cybersecurity has been increasing. 

Machine learning is increasing in popularity to detect risks and automatically eliminate them before they can wreak mayhem. It can filter through millions of files and detect potentially dangerous ones. This was accomplished by Microsoft’s software in early 2018.

According to the firm, hackers utilized Trojan spyware to infiltrate hundreds of thousands of systems and run rogue cryptocurrency miners. Microsoft’s Windows Defender, a software that utilizes many layers of machine learning to identify and block potential threats, effectively blocked this attack. 

As a result, the business was able to shut off the crypto miners as soon as they began digging. Machine learning is used to search for network vulnerabilities and automate actions, in addition to detecting early threats. Machine learning excels at some tasks, such as swiftly scanning vast volumes of data and evaluating it with statistics. Cybersecurity systems create massive amounts of data, so it’s no surprise that this technology is so beneficial. As a result, in the domain of cybersecurity, this is proving to be a big benefit.

Microsoft, Chronicle, Splunk, Sqrrl, BlackBerry, Demisto, and other big corporations are utilizing machine learning to strengthen their cybersecurity systems.

How Modern Data Science Powered by AI Identifies and FIxed IT Vulnerabilities

Here is how data science helps identify and resolve IT vulnerabilities:

1- Improve the Usage of Technologies

Modern Data Science has the potential to both improve and simplify the usage of such technologies. A machine-learning algorithm may be fed both current and historical data through data science. So that the system can detect possible problems accurately over time.

This allows the system to be more precise since it can predict assaults and identify potential vulnerabilities. 

2- Use Encryption

A data breach or assault can cause severe damage to your organization in terms of the loss of important data and information.

This is where data science comes in handy since it uses very sophisticated signatures or encryption to prevent anyone from delving into a dataset. 

3- Create Protocols

Data science has the potential to create impenetrable protocols. By examining the history of your cyber-attacks, you may create algorithms to detect the most often targeted pieces of data. Data science programs may assist you in harnessing the potential of data science to empower networks powered by self-improving algorithms.

Why Should Companies Hire Qualified Professionals?

Thus, the above points indicate the importance of data science and qualified data science professionals in your firm. Focus on hiring professionals who have a master’s degree in engineering in data science and the knowledge of how to decode big data.

We have access to a massive amount of data, and the data is typically telling a narrative. You should be able to identify deviations from the norm if you understand how to analyze data. and such variations can occasionally signal a threat. And, owing to the usage and advancements achieved in machine learning, dangers may now be appropriately countered in a wide range of industries. It is used for image recognition and speech recognition applications.

Even though cybersecurity has improved as a result of this process, humans remain critical. Some individuals believe that you can learn everything from data, but this is just not true. An over-reliance on AI might lead to a false sense of security. 

However, without a doubt, artificial intelligence will become increasingly widespread in maintaining security. It’s maturing, and it’s a feature, not a business. It will play a part in resolving a certain issue. 

Final Thoughts

However, AI cannot address every problem. It will be a tool in the toolbox. At the end of the day, humans are the overlords. 

As a result, in addition to carefully deployed algorithms, cybersecurity specialists, data scientists, and psychologists will play an important role. Human efforts, like those of all existing artificial intelligence and machine learning supplements, augment rather than replace them.

Source Prolead brokers usa