I am sharing my own experience, because, for sure, it will be interesting the same as me, but maybe not only. I will warn you in advance that many terms and abbreviations will be understandable only to those who have basic knowledge and some experience in Data Science and Machine Learning.
So, available for August 2020:
8-year background in e-commerce and targeted advertising
4 courses of Yandex specialization known in the ds-party Machine Learning on the cursor
a couple of courses on neural networks on the "stepik"
slightly disheveled in memory
a couple of python certifications
It was with this baggage that I began to select suitable vacancies and prepare to fight off offers.
1. I studied the resume of my colleagues in search, threw a couple of topics and the resume itself into the career branch in slack Open Data Science (ODS) - I received feedback and the verdict that no one needs such a June. I corrected my resume, it became at least visually pleasant. But the apparent lack of practice hits right in the heart .. Has applied for several internship vacancies. Silence, one refusal.
Nevertheless, light self-promotion does not go in vain. I mentioned somewhere that I analyzed the frequency of mentioning the necessary technologies in the texts of vacancies from the jobs branch in the same ODS slack. A local recruiter asks to drop a resume. There is an nlp vacancy in Gomel. And there are just widespread protests and vintulovo, apparently normal candidates do not want to go ... I would go, "but I have no knowledge in nlp, I can't even pretend". Past. Received recommendations on the content of the resume.
2. That was the time of the ODS datafest and the internship championship from Megafon on the kaggle. This is my chance ...
, , , 3 f1-. : , . , . -10…
3. DS , . , , ODS . .. 2 … ) - , . , .
, , , - , , , ds- . , , . , , . " - )".
4. . 14 100. . . . - 10 "" . . . 14 1.
5. . . . . . 1 . . . , p-value, - . ROC AUC… " 0.5?" - ! … … . : 2 , .
6. . Deep Learning School . : " . , , , . , ?"
, !
" ". . -ML. , "". . , -3 . . beautiful soup , , api, . , , (100 - ).
7. . 48 . 3 . , , . . , / (, , , ). . . … .
, . . . , . . - KNN . ... json , , KNN - . C, - .
. , - . , " ". 5 7 . , .
8. , . , - . , . Json, knn, , , github, catboost. , . . . github. .
9. . . : , , hr. " ". hr. Hr . . - . , nlp. , , . . : " ?". ", . ."
, .
10. . DLS , linux, kubuntu , git. jobs , data .
, . . 3 . , "" , , , - . - .
. 5 . . , " ?". , , . , . " ." )
11. - . ds. . hh. HR. .
. , , , , . . 2 . . 4 - . , .
IT . SQL . . 3 sql. . . . , .
- " ?"
- " , ?"
- " , ? ?"
. " ". . !
data scientist . , . - e M5 kaggle Wallmart. , . . . . eda, . , , , 5 . . .
A couple more invitations are completely past the specialty. Back to Megafon, only the analyst is in Excel. And even the local Vodokanal. And I also joked that I would go as a pipelayer if data science did not work out.
12. At this point, I felt that I needed to take a breath and a week before NG went on vacation to reflect on what had happened. There is no happy ending yet. But the dynamics are positive. I immerse myself in practice, my resume is being completed, I would still pay ...
I think the main thing is to go forward and take your time. All will be.
To be continued…