TimesToCome

Geekiness and life on the third coast

Crash course machine learning




My favorite machine learning resources

If you already have a strong science, math and programming background you can teach yourself machine learning in a relatively short time.

These are the best resources I’ve found so far for learning machine learning:

Basic Data Science, much of the math and deep learning is based on this stuff

Run through the Kaggle tutorial and try solving the Titanic problem using Python with SciKitLearn

Introduction to Statistical Learning

Elements of Statistical Learning

Anaconda Python and all the ML libraries you’ll need to get started

Neural Networks and Deep Learning

Deep Learning using Theano, walks you through building a deep neural network to solve the MNIST problem.

Deep Learning Tutorial from Lisa Labs ( what the Deep Learning with Theano is based on, more of a focus on Theano and less on the networks and math )

Deep Learning Book

Recurrent Networks

WildML Recurrent Neural Networks tutorial

Reinforcement Networks

Deep Reinforcement Learning in 132 lines of Python

Walk through Sutton’s book: Code, links to textbook, videos, papers

Reinforcement Learning, Sutton (textbook)

TensorFlow

TensorFlow and deep learning without a PhD

Machine Learning with TensorFlow ( book ) $

Papers with source code

GitXiv

Papers, all the new important papers are here

Arxiv

Stack Overflow type site for machine learning

CrossValidated

News, tools, data, help

KD Nuggets

Medium Long reads on all sorts of topics, including ML, robotics and AI

Misc

Field Guide to Data Science ( read past the first few pages, it gets better as it progresses )

Beating Kaggle the Easy Way ( lots of good beginner’s tips and tricks )

The Neural Network Zoo

Tips for training deep networks

Theano, a short practical guide

Computer vision for dummies

A Kaggle Master Explains Gradient Boosting

Videos

Google’s Deep Learning Summer School 2016 day 1 lectures, ( long empty intro skip ahead a bit )

Google’s Deep Learning Summer School 2016 day 2

Tutorial, Reinforcement Learning with Sutton

Classics that still hold true

More is Different

The Mathematical Theory of Communication

Let’s Take the Con Out of Economics