/Linear-Models

ICDSS Machine Learning Workshop Series: Linear Models

Primary LanguageJupyter NotebookMIT LicenseMIT

Workshop #1

Prerequisites

  • Basic Linear Algebra
  • Any experience with programming

Overview

The aim of this workshop is to introduce you to Data Science and especially Linear Models. We will answer questions, such as "what is a model?" and "why linear in particular". Then, we will go through some applications, starting with a Simple Beta Hedging algorithm, usually used in Finance. Finally, we will get our hands dirty with implementing this algorithm in vanilla Python and then using off-shelf Machine Learning frameworks, such as scikit-learn and TensorFlow.

Agenda

Theory

Linear Models

Applications

Finance - Simple Beta Hedging

Codelab

Vanilla Python

Setup

  1. Follow Python setup environment
  2. Run make setup command

Resources

Academia

  • Regression Analysis, MIT 18.S096 Topics in Mathematics with Applications in Finance [PDF]
  • The Linear Model I, Caltech CS 156 Machine Learning [PDF]
  • The Linear Model II, Caltech CS 156 Machine Learning [PDF]
  • Linear Regression, Oxford Machine Learning [PDF]

Tutorials

  • Python NumPy Tutorial, Stanford CS231n [tutorial]
  • Linear Regression Example, scikit-learn [code]
  • Linear Regression in TensorFlow, aymericdamien [ipynb]
  • Linear Regression, Quantopian [ipynb]
  • Multiple Linear Regression, Quantopian [ipynb]
  • GradientDescentExample, mattnedrich [Github]

Videos

  • Regression Analysis, MIT 18.S096 Topics in Mathematics with Applications in Finance [YouTube]
  • The Linear Model I, Caltech CS 156 Machine Learning [YouTube]
  • The Linear Model II, Caltech CS 156 Machine Learning [YouTube]
  • Linear Regression, Oxford Machine Learning [YouTube]