Hacker News new | past | comments | ask | show | jobs | submit login

> If you’ve completed a math degree or some other degree that provides an emphasis on quantitative skills, you’re probably wondering if everything you learned to get your degree was necessary.

It should be emphasized that obtaining a degree in a subject like math and physics should "form" your way of thinking in a certain way: That you are able to apply methods to other fields, such as data science. That's way more important then remembering "everything you learned". You might never do a Fourier transformation by hand again, but you know the concept. And you can apply it.




I've met plenty of bootcamp graduates who "know the math" but still can't move from rote recitation to intuitive relation. You can ask for the form of the loss function for "classification" (i.e., the usual cross-entropy) and get it every time. But it is apparent that without a lot of time (which becomes significantly shorter with prior physics/compsci/STEM experience) they can't look at the various tools as building blocks and assemble them into novel machine learning systems. They only understand inductive bias as it pertains to answering trivia about the differences between CNNs and RNNs. Disclaimer to say this is an anecdote but one that is reflected across a number of my colleagues, most of whom I trust not to fall prey to numerous cognitive biases.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: