CS7641-Machine Learning

Machine Learning algorithm

Azure Machine Learning Algorithm Cheat Sheet

Amazon Machine Learning

Google Machine Learning Crash Course

Google AI Education

Note: The goal of Machine Learning is "Generalization".

One meaning of "fold": A pen or enclosure in a field where livestock, esp. sheep, can be kept. (I don't think this meaning has anything to do with cross-validation, though.) Oh, wait, I found it: "consisting of so many parts or facets." So, n-fold cross validation means the data is in n parts. -ML

Supervised Learning

The data points contain X and labeled Y.

The goal is to study many examples and make prediction.

1. discrete value called classification problem
2. continuos value - regression problem

e.g.

• Decision Trees
• Linear regression
• Bag

Bootstrap Aggregating

Bootstrap aggregating Boosting and AdaBoost clearly explained

Unsupervised learning

The goal is to organize the data or to describe its structure.

1. Group data into clusters, K-Means
2. find ways to look at the complex data so that it appears simpler.

Reinforcement learning

The goal is to choose an action in response to each data point. (e.g. robotics or IoT: sensor data/signals to strategy for next action.)

• Q-Learning
  • Reinforcement Q-Learning from Scratch in Python with OpenAI Gym
  • Simple Reinforcement Learning: Q-learning

Reinforcement Learning Coach - Intel AI

OpenAI Baselines

How to evaluate ML Algorithms

Metrics to Evaluate your Machine Learning Algorithm

Metrics To Evaluate Machine Learning Algorithms in Python

Libraries

scikit-learn

• Scientific computing
• Simple and beginner fridenly
• only supervised and unsupervised learning.
• No Deep learning and reinforcement learning algorithm support

TensorFlow - Google

• ecosystem case study
• tensorlayer - Reinforcement Learning Library for Scientists and Engineers
• Tensorforce: a TensorFlow library for applied reinforcement learning - Google

Keras

• Support Tensorflow and Theano backend
• Deep Learning, Neural Networks
• Deep Reinforcement Learning for Keras

PyTorch - Facebook

• A replacement for NumPy to use the power of GPUs and other accelerators.
• An automatic differentiation library that is useful to implement neural networks.
• pytorch-examples

Ray - Fast and Simple Distributed Computing - UC Berkeley RISELab

More Learning and practices

The Best Tools for Reinforcement Learning in Python You Actually Want to Try

kaggle competition

Learn-Data-Sci

Data Science Blogs

Simple Reinforcement learning tutorials, 莫烦Python 中文AI教学

Learning RL

Learning Reinforcement Learning (with Code, Exercises and Solutions)

Reinforcement Learning: An Introduction (2nd Edition)

reinforcement-learning Github most star

reinforcement-learning-an-introduction

Machine Learning, Tom Mitchell

Dataset

paperswithcode