Going Beyond the Intro Model


Found Numer.ai to be quite a interesting project and would like to go beyond the Tutorial.

I have a half decent knowledge of Pandas, Numpy, Statistical techniques in Econometrics etc. but not of Machine Learning Models a la Sci - Kit Learn / Tensor Flow. What would be the logical next step for someone like myself who would like to build their own models to submit?

Thanks for your help.


I’m sort of in the opposite boat to you, in the sense that I don’t have experience with Pandas . . . In terms of building your own models, I got into machine learning by making my own Perceptron and multi -layered perceptron . . . from there I found out about RNN, LSTM and now I’m starting to feel like I will be able to approach Numerai with at least something.

Still not going to make a serious contribution for a while probably.

Hope it helps


Took me seven days to digest that comment - perhaps shows how slow I am [or how much work I have]
Where can I read up on the Perceptron - is [this link a good point](https://machinelearningmastery.com/implement-perceptron-algorithm-scratch-python/) - or am I jumping into the deep end by going this route?