Linear Algebra for Data Science

So here’s a big scary mathematical idea; Linear Algebra. It’s something my wife has studied towards her physics degree, so I know it’s complicated. There are all these funny words associated with it too.

And knowing all that, I was concerned that it’s a part of Data Science.

So yesterday I decided to do a little research, to see how awful it was going to be.

Linear Algebra uses multiple planes

First thing I discovered, Machine Learning is the main culprit. Machine Learning is another part of Data Science that currently spooks me, but I will encounter it at some point. After a little digging, Linear Regression has similarities with a small part of Linear Algebra, and that part features heavily in Machine Learning. So maybe in this case I’m actually somewhat prepared?

Another thing I discovered was that some Data Scientists don’t actually know Linear Algebra before they start the role! There are articles out there advocating that they learn it, which signals to me that it could be skipped. I won’t be skipping it of course, but good to know that I can get along without it, especially if I struggle.

The more different ways of describing Linear Algebra I find, the less scary it seems. It makes sense, because I often like to phrase things myself in many different ways, so having many different explanations for something helps me feel comfortable and understand it better.

I’ve only just begun ot look at Linear Algebra and can just about multiply a matrix by a vector. I don’t yet know how to multiply a matric by a matrix, or how to simplify, but I’ll take that as a good start.

Tell me what you think

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: