Learning Algorithms for Trading

Lesson 1: How Machine Learning is used at a hedge fund

• Overview of use and backtesting
  • Out of sample
  • Roll forward cross validation
• ML methods we will use in this class
  • Linear regression
  • KNN regression
  • (Random Forest regression considering to drop)
  • Q-Learning
• Supervised ML (intent is that the treatment here is light)
  • Use: Regression
  • Use: Classification
  • Model type: Parametric
  • Model type: Instance-based
• Introduce the problem we will focus on in the rest of the class, namely:
  • Example data, will learn on over a particular year (2012)
  • Will test on over the next two years (2013 2014)
  • It will be "easy" data that has obvious patterns
  • You will create trades.txt and run them through your backtester

Lesson 2: Q-Learning and Dyna

• Long/short
• Overview of use and backtesting
  • Out of sample
  • Roll forward cross validation
• Supervised ML (intent is that the treatment here is light)
  • Use: Regression
  • Use: Classification
  • Model type: Parametric
  • Model type: Instance-based
• Reinforcement Learning (light)
  • Use: Find a policy
• Introduce the problem we will focus on in the rest of the class, namely:
  • Example data, will learn on over a particular year (2012)
  • Will test on over the next two years (2013 2014)
  • It will be "easy" data that has obvious patterns
  • You will create trades.txt and run them through your backtester
• Overview: LinReg, KNN, Decision Trees, Q-Learning

Lesson 3: Time series prediction as an ML problem

[note: need to create fake stock data that has embedded patterns]

Lesson 7: Assessing a learning algorithm

• Now that we have two, (linreg & KNN), let's compare them
• RMS error
• Scatterplot predict vs actual
• Corrcoef