Building a career in Data Science
Key points from Pie & AI’s webinar on Building career in Data Science
A few hours ago I attended a YouTube live webinar organized by deeplearning.ai featuring Ayodele Odubela as the speaker.
This post contains the key points I learned from the webinar. In just under an hour Ayodele (eye-ya-deli) explained the entire anatomy of building a data science career. This was one of the best webinars on data science I have attended so far. It is a must-watch if you are trying to get into data science. Since there is no consensus on who a data scientist is, in this blog post whenever I mention data science, I mean data analysis, machine learning, deep learning, etc. Let’s get to what I learned.
So you have done a data science project
Yes, doing projects is important. There are a few things you should be aware though:
- How you made a certain decision e.g. collecting data one way or the other way.
- Why you made that decision e.g. why did you normalize the data? Why you think normalization was needed?
The point is not the above two specific technical questions. The point is the reasoning of your decisions. You need to know the reasoning for every decision you make and you should be able to explain it (to an interviewer e.g.)
Domain Expertise
- You should understand what metrics are and should be able to create new metrics.
- Data science is a mix of coding and domain expertise. So you need to know which methods work in your industry. E.g. healthcare and finance are two very different domains. Depending on which domain you are in you need to learn the specific methods and techniques important related to that domain.
- You need to have a good understanding of Experimental Design. Hypothesis testing and A/B testing are two things you need to be good at.
Portfolio Project
The Portfolio projects must contain every stage of ML development. The primary point behind this is to show the employers:
- skills that you know how to get data from a lot of resources
- that you can communicate with them. You understand different methods of model training, development and evaluation.
To find a project, explore the fields and domains you are interested in.
Consistency
- Practice consistently. This was the most important point mentioned by her in the webinar
- Don’t have much time gap between theoretical study and application
- Get 1% better each day rather than trying to get 20–50% better. It is a compounding-interest thing. 1% makes you grow quickly.
Marketing
- Use Twitter to connect to business owners or finding your community
- Use your blog to share your expertise
- Don’t just write a post on your blog. Provide insight into your thought process.
These are a few of the things I learned. There is a lot more stuff she had talked about. You can watch the full recording on YouTube here:
She just did not stop there, she is even sharing her book for free on building a data science career:
I hope you find my post useful. Make sure you don’t miss watching her webinar. This is one of the best webinars on building a data science career I have come across.
