Machine Learning course in TMS - for fans of artificial intelligence, neural ...

rabiarah

The monologue from the third "Terminator" does not seem like science fiction today, because artificial intelligence algorithms are already used in almost all areas of your life. Do you want to understand the specifics of their work and protect the world from the uprising of machines? Enroll in the Machine Learning course at TMS school ! Your teacher will be Maxim Stepanovich, an expert with 4 years of experience in Data Science and a person who agreed to give a short interview about the future program and ML in general. LinkedIn of the teacher - https://by.linkedin.com/in/maxim-stepanovich-6ab9691aa - Maxim, tell us a few words about yourself - your experience and background. Why did you decide to become a machine learning specialist? - Actually, everything is quite simple. Back in university, I tried different types of programming - I started with creating mobile and web applications. All this seemed template to me, I still hold the same opinion. I always wanted to do something innovative and solve problems with the most non-obvious result, and standard development looked routine. In an effort to escape this routine, I came across neural networks and machine learning. It turns out that fatigue from the classic work of a developer led me to Data Science. - You specified that before ML you were engaged in "classical" programming. What languages ​​and technologies did your previous work involve? - A standard set of web developer tools - Java + Spring, Python with the Flask and Django frameworks. My soul was not in the web, I tried changing languages, but I did not get anything fundamentally new. - Generally speaking, what are Data Science and Machine Learning? Why do we hear about these areas everywhere today? - In the era of widespread digitalization, people are faced with huge amounts of data. Information is everywhere you look. Millions of comments under YouTube videos, thousands of money transfers, hundreds of loan approval factors for potential recipients - all this information needs to be understood and processed. Medicine, for example, does not stand still - every year doctors discover new diseases and begin to use more complex technologies when making diagnoses. The weight of error also increases - if a doctor selects the wrong treatment, the patient will get worse. The same is true for banks, which lose huge amounts of money due to an incorrectly issued loan. The main task of Data Science is to analyze large amounts of data in different ways. The first case of the direction is to replace routine processes with automation tools to improve productivity. The second is to eliminate the factor of human error in information processing. - In what areas are Data Science and Machine Learning algorithms used? Where are they in demand? - It would be better if you asked where they are not in demand.

Machine learning is currently implemented in all areas of life — from education and finance to medicine and marketing. People use smart speakers that run on ML algorithms and make purchases on marketplaces with recommendation systems built on ML algorithms. Even intrusive ads in the browser use Machine Learning tools.

I can share an interesting case with an example of a not-so-obvious application of Data Science. Several years ago, some large supermarket chains came to the conclusion that they should use machine learning to analyze social media marketing service traffic in stores  using CCTV cameras. Companies began collecting data on how customers behave — what routes they take and where they spend more time. Based on this information, experts changed the location of shelves, increasing monthly profits in some stores by 40-50% (without changing the product range!).


— The case is really interesting, but from the outside it looks gloomy. It turns out that Big Brother is watching us, and ML is a kind of evil?

— The question is, frankly speaking, debatable. Any creation can always be described from two sides — positive and negative. By introducing Data Science into medicine, we reduce the risk of medical error, but take work away from doctors.

We should always start from the fact that the Machine Learning ecosystem is managed by a person. Algorithms are implemented in those places where it will be profitable from the point of view of business automation. It is better to buy or write a project tied to AI once than to pay salaries to employees every month.

When Tesla experts developed their famous autopilot, they started talking about the areas of its application. No matter what safety tests the program passed, both the company and the drivers themselves refused to implement it everywhere. No one wants to take legal responsibility if the algorithms suddenly stop working correctly and cause a fatal accident.



— What should a specialist in machine learning know and be able to do? What will you teach students in Machine Learning courses at the TeachMeSkills IT school?

— Data Science is a broad concept with a large set of areas of application. There are countless tasks that can be solved using ML algorithms — I have already given examples from the world of medicine, finance, and marketing. In the course at TMS, we will focus on two areas — NLP (Natural Language Processing — text processing) and computer vision (Computer Vision — image processing).

As a pleasant addition to the presented blocks — an introduction to the MLOps direction. Any models, pipelines, and lines of code written in the field of Data Science are of no value in themselves until they are implemented somewhere. We will also touch on this process — we will get acquainted with Docker, consider implementation approaches, and talk about how ML benefits people.

— What are the advantages of the Machine Learning course at TMS? Will the student be able to become a sought-after specialist?

— Our advantage is in the demand for the skills taught. Text and image processing technologies are essentially considered concepts and can be applied anywhere. NLP and Computer Vision tools go well with the same medicine — they are used to analyze recorded patient complaints and check ultrasound or X-ray images.

The knowledge that we share with students on the course at TMS will be useful in all areas of Data Science. Specific sets of tools and technologies depend on the project that the candidate is applying for when applying for a job.