Design a learning trading agent that uses a Q-learning-based strategy and technical indicators. This is Part 4 in a four-part series of Machine Learning Algorithms:

• Part 1: Implement the Linear Regression Learner and Decision Tree Learner and generate data that works better for one learner than the other.
• Part 2: Implement Random Tree Learner, Bag learner and Insane Learner. Evaluate all the learners implemented in Parts 1 and 2.
• Part 3: Implement the Q-Learning and Dyna-Q solutions to the reinforcement learning problem.
• Part 4 (this repository): Implement a learning trading agent using Q-learning.

• StrategyLearner.py - Main code for a StrategyLearner that uses Q-learning and technical indicators to trade a stock.
• q_learning_trading_example.ipynb - Example of how a StrategyLearner works. This code is also available in the __main__ function of StrategyLearner.py.
• QLearner.py - Code for Q-learning, a reinforcement learning technique. It was first developed and tested in a navigation problem in Part 3.
• indicators.py - Technical indicators to be used as features by a StrategyLearner. They were first developed and used in a rule-based trading strategy
• marketsim.py - A market simulator that computes and visualizes statistics from trades generated by a StrategyLearner. It was first developed in Computational Investing, but has been updated here to read in a dataframe of trades instead of a csv file. There is also an update on how it visualizes a stock's performance.
• analysis.py - Helper code for marketsim.py. It was first developed in Analyze Financial Data
• util.py - Util functions for StrategyLearner.py, e.g. loading data, creating dataframes, etc.
• grade_strategy_learner.py - Main test code for StrategyLearner
• grading.py - Helper test code for grade_strategy_learner.py

You need Python 2.7.x or 3.x (recommended), and the following packages: pandas, numpy, and matplotlib.

Data files can be downloaded from this link or from Yahoo Finance

Place the data into a directory named 'data' and it should be one level above this repository.

To run any script file, use:

python <script.py>

To run any IPython Notebook, use:

jupyter notebook <notebook_name.ipynb>

Source: Part 3 of Machine Learning for Trading by Georgia Tech