MIT xPRO: DSx Data Science and Big Data Analytics: Making Data-Driven Decisions

TOOLS

I. Python

• Beginner (OCW)
• Intermediate (OCW)

II. R

• Beginner (OCW)
• Download R

III. MatLab

• Introduction
• Matlab 30-day trial

RESOURCES BY MODULE

Module 1: Making sense of unstructured data

Instructors: Stefanie Jegelka

• Case Study 1: PCA: Computing Eigenfaces

  • Case Study Package: instructors.zip

• Case Study 2: Spectral Clustering New Stories

Instructors: Tamara Broderick

• Case Study 1: Genetic Codes

• Genetic Code Package.zip

• Case Study 2: Finding Themes in Project Description

Recommended Reading

• Capturing knowledge of user preferences: ontologies in recommender systems

• PCA and K-Clustering of Delta Aircraft

Module 2: Regression and Prediction

Instructor: Victor Chernuzkov

Instructor Slides

• Regression 1.1 - Introduction
• Regression 1.2 - Linear Regression for Prediction
• Regression 1.3 - Assessment of Prediction Quality
• Regression 1.4 - Case Study: Predicting Wages
• Regression 1.5 - The Inference Problem for Linear Regression
• Regression 1.6 - Case Study: Gender Wage Gap
• Regression 1.7 - Other Types of Regression
• Regression 2.1 - Modern Linear Regression for High-Dimensional Data
• Regression 2.2 - High-Dimensional Sparse Models and Lasso
• Regression 2.3 - Inference with Lasso
• Regression 2.4 - Case Study: Do poor countries grow faster than rich countries?
• Regression 2.5 - Other Penalized Regression Methods. Cross-Validation
• Regression 3.1 - Modern Nonlinear Regression. Trees, Random Forests, and Boosted Trees
• Regression 3.2 - Neural Networks
• Regression 3.3 - Assessment of Prediction Quality. Aggregation of Predictors. Case Study
• Regression 3.4 - Inference using Modern Nonlinear Regression Methods. Case Study
• Regression 4.0 - Causality: When It Can and Cannot be Established
• R-Start-up Instructions

Case Study 1: Predicting Wages 1

• Readme document: Module 2 - Predicting Wages 1
• Case Study 1 Package: Module2_CS1_Predicting-wages-1.zip

Case Study 2: Gender Wage Gap

• Readme document: Module 2 - Gender Gap
• Case Study 2 Package: Module2_CS2_gender-wage-gap.zip

Case Study 3: Do poor countries grow faster than rich countries?

• Readme document: Module 2: Rich Countries vs. Poor Countries
• Case Study 3 Package: Module2_CS3_poor-vs-rich-countries.zip

Case Study 4: Predicting Wages 2

• Readme document: Predictive Wages 2
• Case Study 4 Package: Module2_CS1_Predicting-wages-2.zip

Case Study 5: The Effect of Gun Ownership on Homicide Rates

• Readme document: Effect of Gun Ownership
• Case Study 5 Package: Module2_CS1_homicide-rates.zip

Recommended Reading

• Regression in the Olympics
• Regression in Health

Module 3.1: Anomaly Detection and Hypothesis Testing

Instructors: David Gamarnik & Jonathan Kelner

• Case Study: Challenger
• Challenger-data.csv

Recommended Reading

• IBM Fraud Detection

Module 3.2: Deep Learning

Instructors: Ankur Moitra

• Case Study: Decision boundary of a deep neural network

Recommended Reading

• Paper on methods used to create AlphaGo

Module 4: Recommendation Systems

Instructors: Devavrat Shah & Philippe Rigollet

• Case Study 1: Recommending Movies
• Case Study Python Notebook Code Tutorial
• Jupyter Notebook
• Assessment Rubric
• Case Study 2: Recommend New Songs to users based on their listening habits
• Case Study 3: Make new Product Recommendations

Recommended Reading

• Linear Bandits in High Dimension and Recommendation Systems
• Three Challenges for Recommendation Engines

Module 5: Networks and Graphical Models

Instructors: Guy Bresler & Caroline Uhler

Case Study 1: Identifying New Genes that cause Autism

• Case Study Activity Documentation
• Autism-Case-Study-Package.zip

Case Study 2.1: Kalman Filtering: Tracking the 2D Position of an Object when moving with Constant Velocity

• Case Study Activity Documentation
• Kalman-Constant-Velocity-Case-Study-Package.zip

Case Study 2.2: Kalman Filtering: Tracking the 3D Position of an Object falling due to gravity

• Case Study Activity Documentation
• Kalman-3D-Tracking-Case-Study-Package.zip

Recommended Reading

• Social Network Advertising
• Bayesian networks for protein-protein interactions

Module 6: Predictive Modeling for Temporal Data

Instructor: Kalyan Veeramachaneni

Case Study 6.1: 

• Case Study Activity Documentation
• Case-Study-Package.zip
• Rubric

Case Study 6.2:

• Case Study Activity Documentation
• Case Study Data

DISCLAIMER:

These Optional Case Study tutorials will require some prior knowledge and experience with the programming language you choose to use for reproducing case study results. Generally, participants with 6 months of experience using “R” or “Python” should be successful in going through these exercises. MIT is not responsible for errors in these tutorials or in external, publicly available data sets, code, and implementation libraries. As these tutorials are not required (except the two peer-reviewed case studies), we do not offer formal support from the faculty or TAs. However, should you have questions or need assistance, we recommend you utilize the Discussion Forums to pose questions and exchange ideas with your fellow participants. Please note that any links to external, publicly available websites, data sets, code, and implementation libraries are provided as a courtesy for the student. They should not be construed as an endorsement of the content or views of the linked materials.

Student-added resources

Suggest your TAs to add interesting resources in this space (*you can do it from the discussion forum: ASk your TAs sections)

MODULE 2

• Tensorflow Neural Network Simulation shown in class. FKG
• Homepage of Profesor Chernozhukov with a great collection of papers covering many of the topics in the class. FKG
• Analysis p-value histograms. Interesting analysis of how to interpret different p-value histograms.

MODULE 3

• Inspecting Algorithms for Bias, MIT Technology Review magazine (discussion of the risk of bias and metrics for evaluating classifier systems used for criminal justice decisions)