This was a term project for the course CS 2750: Machine Learning (Dr. Milos Hauskrecht, University of Pittsburgh, Department of Computer Science, Spring 2019).

The code tests selected machine learning algorithms on the problem of predicting photo-z's and evaluates the errors as a function of training sample size.

Final report: final_report/Troy-Raen_FinalReport.pdf

Code for ML training and predicting: matlab/

Final plots: data/errors_plots/

Code to make final plots: data/plot_errors.py

Code to process data files: data/data_proc.py

See matlab/pipeline.md for testing process and more plots.