Learning Objectives

• Optimization
• Architecture
• Data Consideration
• Training to Optimize
• ML Considerations (Regularization & Overfitting)

Learning Objectives

• NN vie as Linear Classifiers
• Computation Graphs
• Backpropagation
• Backprop via AutoDiff
• DAG Logistic Regression (Sigmoid)
• Simple Layer Jacobians and Vectorization

Learning Objectives

• What is neural network?
• Supervised Learning
• Parametric Learning
• Performance Measurement
• Linear Algebra Review
• DL v ML differences
• Logistic Regression and Gradient Descent

• Dyna-Q Big Picture
• Learning T
• Learning R
• Dyna-Q Recap

• What is Q
• Learning Procedure
• Update Rule
• Two Finer Points
• The Trading Problem: Actions
• Creating the State
• Discretizing
• Q-Learning Recap

• The RL Problem
• Mapping Trading to RL
• Markov Decision Problems
• Unknown Transitions and Rewards
• What to Optimize
• RL Summary

• Ensemble Learners
• Bootstrap aggregating: Bagging
• Boosting

• A closer look at KNN solutions
• What happens as K varies quiz
• What happens as D varies quiz
• Metric 1: RMS error
• In sample vs. Out of sample
• Cross validation
• Roll forward cross validation
• Metric 2: Correlation
• Overfitting

• Parametric Regression
• K-nearest neighbor
• Kernel regression
• Training and testing
• Learning APIs

• The ML problem
• Supervised Regression Learning
• How it works with stock data
• Backtesting