{"id":30085,"date":"2019-10-31T21:33:35","date_gmt":"2019-10-31T18:33:35","guid":{"rendered":"https:\/\/prohoster.info\/blog\/kak-ya-ne-stal-spetsialistom-po-mashinnomu-obucheniyu\/"},"modified":"2019-10-31T21:33:35","modified_gmt":"2019-10-31T18:33:35","slug":"kak-ya-ne-stal-spetsialistom-po-mashinnomu-obucheniyu","status":"publish","type":"post","link":"https:\/\/prohoster.info\/en\/blog\/kak-ya-ne-stal-spetsialistom-po-mashinnomu-obucheniyu","title":{"rendered":"How I did not become a machine learning specialist.","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"

Everyone loves success stories, and there are plenty of them on Habr.<\/p>\n

<\/p>\n

\u201cHow I got a job with a salary of $300,000 in Silicon Valley\u201d
\n\u201cHow I got a job at Google\u201d
\n\u201cHow I earned $200,000 at 16\u201d
\n\u201cHow I made it to the Top AppStore with a simple currency exchange app\u201d
\n\u201cHow I...\u201d and a thousand and one similar stories.<\/p>\n

<\/p>\n

\"How
\nIt's great that a person achieved success and decided to share it! You read it and feel happy for them. But most of these stories share one thing in common: you can't replicate the author's path! Either you're living at the wrong time, or in the wrong place, or you were born a boy, or... <\/p>\n

<\/p>\n

I believe that stories of failure are often more useful in this regard. You simply don\u2019t need to do what the author did. And that, let\u2019s agree, is much easier than trying to replicate someone else's experience. The problem is that people usually don\u2019t want to share such stories. But I will.<\/p>\n

<\/a><\/noindex><\/p>\n

I worked for many years in systems integration and tech support. A few years ago, I even went to work as a systems engineer in Germany to earn more money. But the field of systems integration no longer inspired me, and I wanted to switch to something more lucrative and interesting. And at the end of 2015, I came across an article on Habr \u201cFrom Physicists to Data Science (From Engines of Science to Office Plankton)\u201d<\/a><\/noindex>, in which Vladimir describes his path into Data Science. I realized: this is what I need. I had a decent knowledge of SQL and was interested in working with data. I was particularly impressed by charts like these:<\/p>\n

<\/p>\n

\"How<\/p>\n

<\/p>\n

Even the minimum salary in this field was higher than any salary I had received in my entire previous life. I firmly decided to become a machine learning engineer. Following Vladimir's example, I enrolled at coursera.org in a specialization of nine courses: \u201cData Science\u201d<\/a><\/noindex>. <\/p>\n

<\/p>\n

I completed one course per month. I was very diligent. In each course, I completed all the assignments until I achieved the highest score. At the same time, I took on challenges on kaggle, and I even succeeded!!! Of course, winning places were out of reach for me, but I ranked in the top 100 several times. <\/p>\n

<\/p>\n

After completing five successful courses on coursera.org and another one titled 'Big Data with Apache Spark' on stepik.ru, I felt empowered. I realized I was starting to grasp the subject. I understood when to use which analytical methods. I became quite adept with Python and its libraries. <\/p>\n

<\/p>\n

My next step was analyzing the job market. I needed to find out what else I should know to secure a job. What subject areas should I study that would interest employers? Alongside the remaining four courses, I wanted to take something more specialized\u2014something that a specific employer would want to see. This would increase my chances of getting a job as a beginner with solid knowledge but no experience. <\/p>\n

<\/p>\n

I went to a job search website to carry out my analysis. However, there were no vacancies within a 10-kilometer radius. Not in a 25-kilometer zone either. Not even in a 50-kilometer radius!!! How could this be? It can't be!!! I went to another website, then another... Then I opened a map of job vacancies and saw something like THIS:<\/p>\n

<\/p>\n

\"How<\/p>\n

<\/p>\n

It turned out that I live right in the center of an abnormal python exclusion zone in Germany. Not a single damn acceptable job for a machine learning specialist or even a Python developer within a 100-kilometer radius!!! This is a fiasco, buddy!!!<\/p>\n

<\/p>\n

\"How<\/p>\n

<\/p>\n

This picture perfectly reflects my state at that moment. It was a below-the-belt blow that I inflicted on myself. And it really hurt...<\/p>\n

<\/p>\n

Yes, I could have gone to Munich, Cologne, or Berlin \u2014 there were job openings there. But one serious obstacle stood in my way.<\/p>\n

<\/p>\n

Our initial plan when moving to Germany was straightforward: go where we are accepted. We didn't care which city in Germany we would end up in. The next step was to settle down, complete all paperwork, and improve our language skills. Then we would head to a big city to earn more. Our preliminary prospective goal was Stuttgart. A large tech city in southern Germany. Not as expensive as Munich. It's warm there and grapes grow. There are many industrial enterprises, which means many job vacancies with good salaries. A high standard of living. Exactly what we need.<\/p>\n

<\/p>\n

\"How<\/p>\n

<\/p>\n

Fate brought us to a small town in the heart of Germany with a population of about 100,000. We got settled, adapted, and took care of all the paperwork. The town turned out to be very cozy, clean, green, and safe. The kids started attending daycare and school. Everything was nearby. There were very friendly people all around. <\/p>\n

<\/p>\n

However, in this fairytale, there were not only no vacancies for machine learning specialists, but even Python was of no use to anyone.<\/p>\n

<\/p>\n

My wife and I began discussing the option of moving to Stuttgart or Frankfurt\u2026 I started looking for vacancies, checking the requirements from employers, while my wife began searching for an apartment, daycare, and school. About a week into the search my wife said to me, \"You know, I don\u2019t want to go to Frankfurt, Stuttgart, or any other big city. I want to stay here.\" <\/p>\n

<\/p>\n

And I realized that I completely agreed with her. I was also tired of living in a big city. It wasn't until I lived in Saint Petersburg that I truly understood this. Yes, a big city is the ideal place to build a career and earn money. But not for a comfortable family life with children. For our family, this small town turned out to be exactly what we needed. It had everything we had been missing in Petersburg. <\/p>\n

<\/p>\n

\"How<\/p>\n

<\/p>\n

We decided to stay until our children grew older. <\/p>\n

<\/p>\n

But what about Python and machine learning? And the six months I have already spent on all of this? Well, it doesn\u2019t matter. There are no job vacancies nearby! I didn't want to spend 3-4 hours a day commuting to work anymore. I had already done that in Petersburg for several years: commuting from Dybenko to Krasnoye Selo before the ring road was built. An hour and a half there and an hour and a half back. Life passes by, and you watch the flickering houses from the car or minibus window. Yes, you can read, listen to audiobooks, and all that. But it quickly becomes tiresome, and after six months to a year, you find yourself just wasting time listening to the radio, music, and aimlessly staring into the distance.<\/p>\n

<\/p>\n

I had setbacks before. But I hadn\u2019t made such foolish mistakes in a long time. The realization that I couldn\u2019t find a job as a machine learning engineer threw me off balance. I dropped all my courses. I stopped doing anything at all. In the evenings, I drank beer or wine, ate salami, and played LoL. That\u2019s how a month went by. <\/p>\n

<\/p>\n

In reality, it doesn't matter what difficulties life throws at you. Or even what you impose on yourself. What matters is how you overcome them and what lessons you extract from those situations. <\/p>\n

<\/p>\n

\"What doesn't kill us makes us stronger.\"<\/strong> You know this wise phrase, right? Well, I believe it's complete nonsense! I have a friend who lost his job as a director of a fairly large car dealership in St. Petersburg during the 2008 crisis. What did he do? That's right! Like a real man, he went out to look for a job. A director job. And when he couldn't find a director position for six months? He continued to look for a director job, but in other fields, because working as a sales manager for cars or anything other than a director was beneath him. In the end, he found nothing in a year. And then he just gave up on the job search altogether. His resume still hangs on HH \u2013 if someone needs him, they will call. <\/p>\n

<\/p>\n

And he was unemployed for four years while his wife earned money the whole time. After a year, she got a promotion, and they had more money. Meanwhile, he just sat at home, drank beer, watched TV, and played video games. Of course, it wasn't just that. He cooked, did laundry, cleaned, and went shopping. He turned into a bloated boar. Did all of this make him stronger? I don't think so.<\/p>\n

<\/p>\n

I could have also continued drinking beer and blaming employers for not posting job openings in my village. Or blame myself for being such a fool and not even bothering to look for vacancies before diving into Python. But there was no point in that. I needed a Plan B\u2026<\/strong><\/p>\n

<\/p>\n

In the end, I gathered my thoughts and started with what I should have begun with in the first place \u2013 with market analysis. I analyzed the job market in the IT field in my city and concluded that nearby there are:<\/p>\n

<\/p>\n