5 Rules of getting a job in data science

5 Rules of getting a job in data science

This is inspired by Jordan Peterson's practical advice book "12 rules for life"

First - Know the rules

Understand the profile and get your keywords right. The search engine uses cosine distance on TF-IDF kind of scheme. JD-CV keyword overlap is extremely crucial.

Having rare keywords will get a higher weightage to your resume because IDF value will be high. (The actual scheme is most probably BM25 which is a modified version of TF-IDF.)

How do I know all this stuff?

I have worked and talked with people who do data science in HR.

As a fresher I had to do resume rewriting and keyword gaming so that I would get shortlisted. It's a tough game and you need to do whatever it takes. (Don't fake anything. It can spoil reputation forever.)

Nowadays I do the reverse - Shortlist jobs as per my goal.

Second - Hard truth

If your resume is already the best you can have, you applied for 100 jobs and you still didn't get interview, just leave job application on the side.

This means that your profile is not yet ready or isn't above average. Swallow the hard truth.

Job application can look like a 5-minute button click activity but god knows how much time it can consume. So just stop wasting time.

Third - You and your courses

There is a flood of courses - free and paid.

Most of the time, the quality of course is inversely proportional to its price. This also means that a free course has infinite quality - it's actually true.

The price of some courses isn't high because they are of high quality. Its because a lot of their money goes into advertising and failure to do economies of scale.

Default to Coursera. Audit a course if you have no money and find the course assignments on GitHub.

My favorite nowadays is YouTube.

Fourth - Certificate validation game

I have been working in AI for a few years and it struck me one day that nobody ever asked me for my certificates. Seems, the industrial value of my certifications is zero.

Rather than chasing certificates, find your USP(Unique selling point) or the Ahaa factor - someone should look at your profile and get immediate Ahaa to hire you.

I leave this to you to define your own way to raise your USP.

Think…

What am I supplying
that no one else can supply?


Fifth - Inner fears

Every moment you spend reading How to get in data science? you are just wasting time.

Stop spending time reading articles like this.

This is what we call procrastination.

There is a high chance that you already know enough.

Start writing code. Be consistent. Show up every day.

Tapan Patro

Senior Data Scientist @ Tide

4 年

Can you share some point on how to overcome section: Second - Hard truth. if the profile is not ready implies working on realtime applications? if yes... I have done my end to end project, with my own data collection to deployment https://medium.com/analytics-vidhya/end-to-end-deep-learning-based-app-af67d4008550 . How many of these projects could make my profile good ?

回复
Yogesh Kothiya

Data Analytics | AI | Data Science | Mentor | Content Creator | Community Builder

4 年

Free course has infinite quality - it's actually true Damn Yes! A lot of their money goes into advertising. Hell lot

Rahaman Sheriff

Doctorate Scholar | Principal Product Manager - Order Management | NIT Warangal | Amity

4 年

Pratik Bhavsar I totally agree with ur 5th point. That suffice the quote one punch at a time.

要查看或添加评论,请登录

Pratik Bhavsar的更多文章

  • Mental Models In Data Science

    Mental Models In Data Science

    Slow-thinking vs fast-thinking From Google’s 43 rules of ML “Rule #4: Keep the first model simple and get the…

  • On Variety Of Encoding Text

    On Variety Of Encoding Text

    Encoding text is at the heart of understanding language. If we know how to represent words, sentences and paragraphs…

    4 条评论
  • On Semantic Search

    On Semantic Search

    It took me a long time to realise that search is the biggest problem in NLP. Just look at Google, Amazon and Bing.

    4 条评论
  • How to train your Neural Network

    How to train your Neural Network

    The value of a neural network lies in its hypertuning. General intuition The VA(validation accuracy) of your NN(Neural…

    6 条评论

社区洞察

其他会员也浏览了