Data Science for Finance

Introduction

This course covers the fundamental statistical tools used by quantitative analysts with a focus on forecasting financial series. Several key methods used in modern data science (also known as “statistical inference” and “machine learning”) are presented and applied to financial data. The textbooks are widely adopted in machine learning courses but they do not focus on financial data; the necessary adaptations will be discussed in class and in the lecture notes. We will also learn how to move past the conditional mean and model other moments of the conditional distribution, such as quantiles. 

Course content

• Introduction to statistical learning. Features of financial data.
• K-nearest neighbour. The bias-variance trade-off.
• Least squares and linear projections. Weighted least squares. Local regression.
• Penalization: ridge regression and lasso.
• Cross-validation.
• Modelling nonlinearities with regression splines.
• Generalized linear models and maximum likelihood.
• Penalized ML. Weighted ML. Local ML. Splines and ML.
• Regression trees and boosting.
• Nonlinear modelling of variance and quantiles.
• Model stacking.  
• Introduction to Bayesian statistical thinking (if time allows).

Disclaimer

This is an excerpt from the complete course description for the course. If you are an active student at BI, you can find the complete course descriptions with information on eg. learning goals, learning process, curriculum and exam at portal.bi.no. We reserve the right to make changes to this description.