8 ms·
DON'T LEARN NEURAL NETWORKS FIRST. Instead, learn decision trees and more importantly enough statistics so you aren't dangerous. Do you know what the central
by nl 10y ago
DON'T LEARN NEURAL NETWORKS FIRST.
Instead, learn decision trees and more importantly enough statistics so you aren't dangerous.
Do you know what the central limit theorem is and why it is important? Can you do 5-fold cross validation on a random forest model in your choice of tool?
Fine, now you are ready to do deep learning stuff.
The reason I say not to do neural networks first is because they aren't very effective with small amounts of data. When you are starting out you want to be able to iterate quickly and learn, not wait for hours for a NN to train and then be unsure why it isn't working.
- scottmcdot 10y agoThanks. Can you recommend any statistics books to be safe?
- hnarayanan 10y agoI found this a really good book: http://www-bcf.usc.edu/~gareth/ISL/ http://www-bcf.usc.edu/~gareth/ISL/
- master_yoda_1 10y agohttp://www-bcf.usc.edu/~gareth/ISL/ http://www-bcf.usc.edu/~gareth/ISL/ is not an statistics book. its a statistical machine learning book. "All of statistics" is really a great book if you have time work through he exercise.
- stared 10y agoMacKay, "Information Theory, Inference, and Learning Algorithms" and taking the Bayesian Inference and Machine Learning path http://www.inference.phy.cam.ac.uk/itila/book.html http://www.inference.phy.cam.ac.uk/itila/book.html (freely accessible online)
- adamnemecek 10y ago"Introduction to probability" is amazing. It really set the bar for what an academic book could be https://www.amazon.com/Introduction-Probability-Chapman-Statistical-Science/dp/1466575573 https://www.amazon.com/Introduction-Probability-Chapman-Stat...
- mindcrime 10y agoThere are some good, short, MOOC courses on statistics and probability on Coursera these days. I've been working my way through the Duke sequence with Mine Çetinkaya-Rundel and have found them very helpful. The courses correspond with the material in this OpenIntro text: https://www.openintro.org/stat/textbook.php https://www.openintro.org/stat/textbook.php
- rsp1984 10y agoI don't think it's a good strategy to discourage people from diving right in. There are many courses and books out there that are suitable even for a beginner who wants to learn about NN. Of course it's important to get a broad horizon eventually but starting with the theory without the applications is not how most humans learn best. Learning by doing is.
- nl 10y agoI agree strongly with the idea of diving in. The problem with diving into neural networks is that they are slow to train (with large amounts of data anyway), and difficult to debug. This means it isn't really a great place to start.