Anáhuac Machine Learning

These are materials for a machine learning course at Anahuac, February 5 - 17, 2020. If you have any questions or comments please email me at hersh [at] chapman [dot] edu

Syllabus: Anahuac_Machine_Learning.docx

---

Schedule

• Class 1: February 5 — 13:00 - 16:00 — SDCAIDE3
  1. Introduction
  2. Data Visualizations and Rmarkdown
     • Class1-1_Intro_DataViz.r
     • In-class: Feb5_firsthalf.r
  3. Introduction to R, Basic Data Transformations and Exploratory Data Analysis
     • Class1-2_RBasics_Exploratory_Data_Analysis.r
     • In-class: Feb5_firsthalf.r
• Slides: Class_1_Anahuac_ML_Introduction.pptx
• Homework for Friday: Read R for Data Science Chapters 1-7

• February 7, 13:00-16:00, SCAD6
  1. More exploratory data analysis: group_by(), arrange()
  2. Bias-Variance Tradeoff
  3. Linear Models in R

• Slides: Class_2_Anahuac_ML_Bias_Variance_OLS.pptx
• Code: class_2_Test_Training_Split_Bias_Variance.r
• In-class code: Feb7_inclass.r
• Homework: read ISLR Chapters 1-2, Problem Set 1 link: Anahuac_ML_pset1.pdf

• February 10, 13:00-16:00, SDCAIDE1 - Classification
  • Logistic Regression
  • Interpreting logistic regression and estimation in R
  • Classification diagnostics: ROC Curves, AUC, calibration,Confusion matrices, false/true positives and negatives, lift charts
  • Severe class imbalance
• Slides: Class_3_Classification.pptx
• Code: class_3_classification.r
• In-class code: Feb10_inclass.r

• February 12, 13:00-16:00, SDCAIDE1 - Regularized Regression (Lasso, Ridge, ElasticNet)
  • Cross-validation
  • Forward and backward stepwise selection
  • Ridge
  • Lasso
• Slides: Class_4_Cross_Validation_Regularization.pptx
• Code: class_4_Cross_Validation_Ridge_Lasso.r

• February 14, 13:00-16:00, SDCAIDE1 - ElasticNet, Decision Trees and Random Forest
  • ElasticNet
  • Decision Trees
  • Random Forests
• Slides: Class_5_ElasticNet_Tree_Methods.pptx
• Code: class_5_ElasticNet_Trees.r

• February 17, 13:00-16:00, SDCAIDE1 - Unsupervised learning
  • K-Means clustering
  • Hierarchical Clustering
  • Principal Component Analysis (PCA)
• Slides: Class_6_Unsupervised.pptx
• Code: class_6_Unsupervised.r
• Fun with R - gganimate:
  • Fun_gganimate.r
  • https://github.com/thomasp85/gganimate

---

Textbooks

• R for Data Science https://r4ds.had.co.nz/
• James, G., Witten, D., Hastie, T., & Tibshirani, R. (2013). An introduction to statistical learning: with applications in R(Vol. 103). Springer Science & Business Media.
• Varian, Hal R. “Big data: New tricks for econometrics.” Journal of Economic Perspectives 28, no. 2 (2014): 3-28.
• Kuhn, M., & Johnson, K. (2013). Applied predictive modeling. Springer Science & Business Media.

---

Problem Sets

• Problem Set 1
  • Anahuac_ML_pset1.pdf
• Problem Set 2
  • Anahuac_ML_pset2.pdf

Datasets

• Movie Metadata
  • Link: https://jonathan-hersh.com/wp-content/uploads/2020/02/movie_metadata.csv
  • Metadata: https://jonathan-hersh.com/wp-content/uploads/2020/02/movie_metadata.txt

• Housing Prices
  • Link: https://jonathan-hersh.com/wp-content/uploads/2020/02/housing_prices.csv
  • Metadata: https://jonathan-hersh.com/wp-content/uploads/2020/02/housing_prices_data_description.txt

• Wholesale Customers
  • Link: Wholesale_customers.csv
  • Metadata: https://archive.ics.uci.edu/ml/datasets/wholesale+customers

• Bike Sharing Usage
  • Link: day.csv
  • Metadata:
    https://www.kaggle.com/c/bike-sharing-demand

---

Additional References:

• Afzal, M., Hersh, J., Newhouse, D. “Building a better model: Variable selection to predict poverty in Pakistan and Sri Lanka” (2015). Working Paper
• Athey, S. (2017). Beyond prediction: Using big data for policy problems. Science, 355(6324), 483-485.
• Athey, S., Imbens, GW. “Machine Learning Methods for Causal effects”.  http://www.nasonline.org/programs/sackler-colloquia/documents/athey.pdf
• Monica Andini, Emanuele Ciani, Guido de Blasio, Alessio D’Ignazio. “Effective policy targeting with machine learning“
• Belloni, A., & Chernozhukov, V. (2009). Least squares after model selection in high-dimensional sparse models.
• Celiku, B., & Kraay, A. (2017). Predicting conflict. The World Bank
• Blumenstock, J. E. (2016). Fighting poverty with data. Science, 353(6301), 753-754.
• Diamond, Alexis; Gill, Michael; Rebolledo Dellepiane, Miguel Angel; Skoufias, Emmanuel; Vinha, Katja; Xu, Yiqing. 2016. Estimating poverty rates in target populations : an assessment of the simple poverty scorecard and alternative approaches. Policy Research working paper; no. WPS 7793. Washington, D.C. : World Bank Group. http://documents.worldbank.org/curated/en/801751471268674333/Estimating-poverty-rates-in-target-populations-an-assessment-of-the-simple-poverty-scorecard-and-alternative-approaches
• Einav, L., & Levin, J. (2014). Economics in the age of big data. Science, 346(6210), 1243089.
• Jean, N., Burke, M., Xie, M., Davis, W. M., Lobell, D. B., & Ermon, S. (2016). Combining satellite imagery and machine learning to predict poverty. Science, 353(6301), 790-794.
• Engstrom, R., Hersh, J., & Newhouse, D. (2016). Poverty from Space: Using high resolution satellite imagery for estimating economic well-being.
• Hersh, J., & Harding, M. (2018). Big Data in economics. IZA World of Labor, 451-451.
• Harding, Matthew & Lovenheim, Michael, 2017. “The effect of prices on nutrition: Comparing the impact of product- and nutrient-specific taxes,” Journal of Health Economics, Elsevier, vol. 53(C), pages 53-71.
• Kleinberg, J., Ludwig, J., Mullainathan, S., & Obermeyer, Z. (2015). Prediction policy problems. The American economic review, 105(5), 491-495.
• McBride, L., & Nichols, A. (2016). Retooling poverty targeting using out-of-sample validation and machine learning. The World Bank Economic Review, 32(3), 531-550.
• Mullainathan, S., & Spiess, J. (2017). Machine learning: an applied econometric approach. Journal of Economic Perspectives, 31(2), 87-106.
• Wager, S., & Athey, S. (2017). Estimation and inference of heterogeneous treatment effects using random forests. Journal of the American Statistical Association, (just-accepted).