What you will learn
1. Core Concepts of ML
2. key ML algorithms that are used in finance
3. Analyze both "classical" ML approaches (logistic regression, SVM, decision trees, etc.) and neural architectures including deep learning approaches
4. Use ML Methods to solve real-world problems in finance using Python ML libraries such as scikit-learn and TensorFlow
5. Complete programming assignments involving examples from equity research, trading, asset management, credit risk management, banking, etc.
Pre-requisites
Python programming skills
- Numpy, Pandas, and iPython/Jupyter notebooks
Linear Algebra (Chapters 2 and 3 in Goodfellow et. al., "Deep Learning" (2016)
- linear matrix equations, eigenvalue decomposition, inverse matrices, and other related concepts
Basic probability theory
- Gaussian, exponential, or binomial distributions, basic probability rules, such as the base landmark, and some basic statistics
Calculus
- particular rules of differentiation of composite functions
In this course, we will:
1. Analyze one of the most classical problems of Quantitative Finance, namely, the problem of pricing financial options
2. Study Modification of the Black-Scholes model to a discrete time formulation
3. Interpret Markov Decision Process as a reformulation of BSM model with a certain reward function
4. Use Dynamic Programming approach to solving the MDP problem
5. Learn some of the main algorithms of Reinforcement Learning, namely, Q-learning and fitted Q iteration
6. Study applications of Reinforcement learning for dynamic management of stock portfolios, especially, Markowitz optimal investment portfolio, and optimal stock trading
7. Explain methods such as inverse reinforcement learning and g-learning which are needed to deal with high dimensional state and action spaces in quantitative finance problems
8. Apply these methods in the final project of this course to learn optimal trading
