Machine Learning for Development, July 2018

These are materials for a machine learning course for the Poverty Global Practice at the World Bank, July 12-13. If you have any questions or comments please email me at hersh [at] chapman [dot] edu


July 12, 2018 

July 12th (9am – 12:15; 2:00 – 5pm)

  1. Introduction and Cross-validation
    • [Slides]; [code]; [Introduction to Statistical Learning: Chp 2 & 5.1]
  2. Shrinkage methods (Lasso and Ridge)
  3. Classification
    • [Slides]; [code]; [APM: Chp 11, 12 & 16, ISLR: Chp 4]
  4. Tree-based methods (Decision trees, bagging, random forests, boosting)
  5. Unsupervised learning (PCA, clustering)

Schedule:
9:00 – 10:30: Lecture
10:30 – 11:00: Coffee break
11:00 – 12:15: Lecture
12:15 – 2:00: Lunch break
2:00 – 3:00: Lecture
3:00 – 3:30: Coffee break
3:30 – 5:00: Lecture

July 13th

9am – noon: Coding examples of material on day 1

Presentations

1:30 – 2:30: Aart Kraay, “Predicting Conflict” (joint with Bledi Celiku)

2:30 – 3:00: Kristen Himelein, “TBD”

3:00 – 3:20: Break

3:20 – 4:20: Leonardo Lucchetti “What can we (machine) learn about welfare dynamics from cross-sectional data?”

4:20 – 5:00: Jonathan Hersh, “Asking the Right Question: Survey Design Using Group Lasso and Sparse Group Lasso” (joint with Leonardo Lucchetti and Nancy Nugent); “Poverty Mapping Using Convolutional Neural Networks Trained on High and Medium Resolution Satellite Images, With an Application in Mexico” (joint with Boris Babenko, David Newhouse, Anusha Ramakrishnan, Tom Swartz)

 

References:

References:

Main texts:

Papers: