Interesting AI/ML Articles You Should Read This Week (July 11)

Analysis and Opinion

Interesting AI/ML Articles You Should Read This Week (July 11)

Discover articles that provide advice and tips on navigating a successful career in Data Science and Machine Learning.

I love the analogies Ganes Kesari uses to describe the struggles of a data scientist, I mean, comparing interaction with Project Managers to getting a tooth removed is just comical (and accurate?).

I’ve seen many individuals ask “H ow can I become a data scientist? ” or “ What programming languages do I need to learn ?”. These questions and many more similar questions are answered in this week selection of interesting articles in AI/ML you should read.

This week’s edition is very relevant for individuals starting in Data Science, or individuals who are seeking to navigate a successful machine learning career.

  • Discover advice from Miguel Pinto ’s journey on how he became a Kaggle competition master.
  • Or you can get some superpowers from Ganes Kesari to help with your Data Science career.
  • How about a scare in regards to the security of your Data Science role. in the well-written article by Frederik Bussler .
  • And an explanation of the Machine Learning Engineering Role by Caleb Kaiser .

Cover images of the articles included

My 3-year journey: From zero Python to Deep Learning competition master By Miguel Pinto

Miguel Pinto lays out his path to becoming a Kaggle competition grandmaster for his readers. His path details his two-year journey in which he learnt Python, Deep Learning and many more related subject areas.

Miguel wrote this article to inspire individuals to take upon data science and all its related aspect. Miguel’s story is motivational as it shows the ability for individuals to progress from little or basic knowledge within a field, all the way to attaining an industry recognised achievement.

The introduction of the article starts off mentioning reasons as to why the popularity of AI has grown. Miguel shows the versatility and transferability of gaining skills in AI by showcasing a list of industries that benefits from the utilisation of AI-based solutions. But Miguel doesn’t shy away from stating the negative impacts that misuse of AI can have.

The recollection of his journey starts with an account of how he set out to learn Python and the difficulties he faced.

I like his no-nonsense and direct approach to providing actionable advice to readers. He has even highlighted his top tips in bold, so you will not miss any.

Taking a course by Andrew Ng is almost a rite of passage for the majority of machine learning students, including Miguel. Miguel includes a reference of what deep learning courses proved the most effective for him.

He also covers how in retrospect he would have approached his Deep Learning studies, which you, the reader can adopt now and probably reach a level of deep learning proficiency in a shorter timeframe.

The “Competing in Kaggle” section of Miguel’s article is probably the reason why most people would read the article. This section contains Miguel’s recollection of his rather quick progression through the Kaggle ranks and the struggles he faced when learning. Again, Miguel has highlighted key points and tips in bold, and these tips are career-defining.

The article ends with a nudge to the benefit of attaining knowledge in Deep Learning, even if you don’t choose it as a career choice.

This article is a great read for:

  • Deep Learning Students
  • Kaggle Platftoms Users

4 Superpowers That Will Make You Indispensable In a Data Science Career By Ganes Kesari

You’ve read Miguel’s article and story above, and you are ready to take on a career in Data Science(or at least curious about one). Well, it’s only right you equip yourself with superpowers.

The introduction of Ganes Kesari ’s article is very vivid and creatively written. I love the analogies Ganes uses to describe the struggles of a data scientist. The comparisons of dealing and interaction with Project Managers to getting a tooth removed is just comical.

Ganes has to paint this rather scary work life of a Data Scientist to present the four superpowers you will need to preserves in the face of villainous adversity( I’m keeping with the superpower theme ).

But on a serious note, Ganes has presented four areas of challenges that a commonly faced by Data Scientist and has provided with tips and advice to help you navigate a career in data science.

Ganes includes tips on how to handle messy data and the importance of mastery. But one key point that struck a chord with me is the emphasis on the importance of focusing on techniques rather than tools and application. I’ll keep it simple, tools and applications come and go, but techniques stay the same.

It’s paramount that Data Scientist can understand the fundamentals of techniques as opposed to the utilisation of application that will be outdated in a couple of years.

All the superpowers that Ganes has equipped the reader with are applicable regardless of what projects you are working on.

Give this article a read if you are currently stuck in a rut at work and need some advice.

This article is an excellent read for

  • Data Science Practitioners

Will AutoML Be the End of Data Scientists? By Frederik Bussler

So Ganes equipped you with some superpowers, but wait, you might not even get a chance to use them!

AutoML has been a hot topic within the machine learning industry for a few years. And if you are unfamiliar with AutoML, don’t worry. Frederik Bussler has put together an article that introduces the promise of AutoML and companies, startup and organisation that are leveraging the approach of automation within machine learning processes.

Frederik writes about the growing popularity of AutoML and its adoption by big tech giants such as Amazon and Facebook, along with several other startups.

The list of companies adopting AutoML as a business shows that this method of automation is only going to become even more prominent in the machine learning world.

Frederik goes on to point a distinction between AutoML and No-Code AI. Simpy kept, No-Code AI is the abstraction of the complexities that can be introduced by heavily coded AutoML solutions.

The rest of the article includes the advantages and disadvantages of AutoML.

A more exciting section of the article of the reader will be the part that focuses on the shortcomings of AutoML( spoiler alert : data scientists will still have jobs ). The shortcoming point to the lack of human intuition and capability in the absence of explainability and also non-bias data gathering.

This article is a great read for:

  • Data Scientists
  • Machine Learning Engineers

Moving from data science to machine learning engineering By Caleb Kaiser

Frederik article might have given you a scare, and now you aspire to dive deeper into the world of AI and become a full-fledged Machine Learning Engineer.

Caleb Kaiser ’s article starts within pointing out the feats that models implemented by machine learning practitioners have achieved; achievements such as human text generation, object detection and text-speech models. Caleb recognises the marrying of the Data Scientist and Machine Learning Engineer role to make many machine learning achievements possible.

Nonetheless, Caleb has tried to create a distinction between ML Engineers and Data Scientists by stating explicitly that: an ML Engineer role is associated with how machine learning can be implemented to solve actual problems.

To create an even clearer distinction, Caleb has included four machine learning application and where the roles of ML Engineers and Data Scientist come into play. These clear distinction would aid Data Scientists to get an idea of potential responsibilities if a career transition were to be made.

I have to admit the lines between Data Scientists and Machine Learning Engineers are blurred at times, mainly since they have shared responsibilities, and it doesn’t help that the duties associated with the roles changes depending on companies.

A good portion of the article is focused on ensuring the distinction between the two roles is clear to the reader; once this is achieved, Caleb goes on to the main body of the article.

The main body of the article focuses on what a machine learning role entails, Calebs employs several application examples to illustrate both the simplicity and complexity of the ML Engineering role.

The closing half of the article points to the application of principles adopted from software engineering into machine learning processes.

This article is a great read for:

  • Data Scientist that want a high-level understanding of what an ML Engineering role entails
  • Machine Learning Engineers that would like to understand how to improve existing processes

I hope you found the article useful.

To connect with me or find more content similar to this article, do the following:

  1. Subscribe to my YouTube channel for video contents coming soon here
  2. Follow me on Medium
  3. Connect and reach me on LinkedIn