- Good knowledge of the fundamentals of Statistics (i.e. descriptive statistics, probability, inferential statistics, linear regression model).
- Previous programming experience is desirable but not requested.
The course aims at providing the knowledge of cutting-edge AI and machine learning (ML) tools for modeling financial data defined in high-dimensional spaces and characterised by non-linear relationships. In particular, the objective of the considered methods is the automatic detection of patterns in the data (i.e. to “learn” from data) by taking into account the specific peculiarities of financial data. The estimated models can then be used by the analysts and investors to take decisions and choose investment strategies under uncertain and risky conditions.
At the end of the course the student will gain the ability to:
a) understand and explain the main machine learning and deep learning techniques;
b) choose and apply the appropriate modeling tool, in the class of ML methods, for the analysis of financial data;
c) use free the open-source software R and/or Python for performing data analysis and visualization and implementing ML models;
d) assess the performance of the implemented predictive methods and interpret all the available results in a decision making perspective.
- Introduction to AI and machine learning: supervised, unsupervised and reinforcement learning, deep-learning, regression and classification problems, the bias-variance trade-off.
- Illustration of financial applications with machine learning methods (e.g. price prediction, portfolio construction, risk analysis, credit ratings, outlier detection, algorithmic trading).
- Training, validation and testing, cross-validation back-testing, hyper-parameter tuning
- Ensemble methods: classification and regression trees, bagging, random forest, boosting.
- Neural networks: feedforward convolutional and recurrent neural network.
- Elements of reinforcement learning.
The course consists of theory lectures and R/Python lab sessions (usually R labs represent 1/4 of the total number of hours).
The exam consists in:
- a test including open-ended and T/F questions concerning theoretical topics or short applications of the studied methods;
- exercises to be solved using the R/Python software in order to evaluate the ability of the student in analysing data and interpreting outputs.
The two parts of the exam (theoretical and practical) are each worth 50% of the total score, approximately. The final scores will be published in the e-learning page of the course.
- Attending lectures and R labs is strongly recommended.
- If the course will be delivered remotely (totally or partially), changes may occur in the program and/or in the exam, in order to adapt the course to on-line teaching methods.