What are all the software skills needed to become a data scientist?

Data Science has become the new desirable IT job. While there are only a few in the market conversant with the terms like python, machine learning, deep learning and transflow, it is also a fact that these skills are high in demand.

0 Upvotes 0 Upvotes 0 Downvotes 1 Comment

• • Report Abuse
• Comment

#1 Programming As others have mentioned, programming is the number one skill. All data Science applications are coded. Not coding limits the options to a degree where I personally cannot consider it to be data Science. In the expression data Science, programming is the data part. Knowing how to handle any body of data implies programming. #2 Math, Stats and Probability Ive written this so many times that I should tattoo this on me (but I won't) or risk getting a warning from Quora for not quoting my own previous posts: Stats and Probability, the more the merrier. In the expression data Science, math, stats and probability are the Science part. This can be many things but to me the most significant part is the design and analysis of experiments which allows to assess the data products we create. #3 Machine Learning I isolated machine learning from the previous point because the process of creating a machine learning based data product is a world in its own. Knowing programming and knowing statistics, even if one knows how to fit statistical models is not sufficient to create such a data product. It is true that at its core machine learning algorithms are not fundamentally different from many of the algorithms used in statistics. However the application differs, namely the concerns with performance, especially in real time applications and feature engineering.

0 Upvotes 0 Upvotes 1 Downvote

• • Report Abuse
• Comment