In this project, we applied model selection methods on different regression models introduced in Chapter 6: Linear Model Selection and Regularization (from "An Introduction to Statistical Learning" by James et al). The 5 regression models we practiced are Ordinary Least Squares, Ridge Regression, Lasso Regression, Principle Components Regression and Partial Least Square Regression. In order to find the best parameter (or coefficients) for those regression models, we splitted our datasets into train and test groups, performed 10-fold cross validation on each model respectively, and in the end chose the best model with the least cross-validation error.

Contents are separated into five main sections (with each section corresponding to a directory)

• Code -- this is where regression scripts & utility functions & unit tests locate.
• Data -- data downloaded from the internet & created by code are stored in this directory.
• Images -- images created by exploratory analysis & regression are stored in this direcotry.
• Report -- contains a section sub-directory that includes all sections of the final report in separate files and final report is dynamically generated and stored here.
• Slides -- slides are dynamically generated and stored here.

And the file strucutre of this project is like the following

Stat159-Project-2/
	.gitignore
	README.md
	Makefile
	LICENSE
	report/
		sections/
			00-abstract.Rmd
			01-introduction.Rmd
			02-data.Rmd
			03-methods.Rmd
			04-analysis.Rmd
			05-results.Rmd
			06-conclusions.Rmd
		report.Rmd
		report.pdf
	images/
		... (dynamically generated images)
	data/
	    Credit.csv
	    ... (dynamically generated data)
	code/
	    functions/
	        ... (utility functions)
	    scripts/
	        ... (regression scripts)
	    tests/
	        ... (unit tests against utility functions)
	 slides/
	    ...
	 session-info.txt (system info)

This project can be reproduced by following the instructions below.

1. Download/Clone this project from GitHub (unzip if downloaded file is in zip format)
2. Open terminal or any shell program that supports standard linux commands
3. cd Stat159-Project-2
4. make clean to remove old compiled artifact
5. make to generate new artifact
6. open report.pdf with PDF viewer of your choice

This section includes description of different make commands that you can use to reproduce corresponding part of this project

1. make all reproduce the entire projects -- download data, run regression analysis, aseemble report etc
2. make data downlaod data from internet, run data cleaning script, split data into train set and test set
3. make eda run exploratory data analysis script
4. make ols run OLS regression script and save result
5. make ridge run Ridge regression script and save result
6. make lasso run Lasso regression script and save result
7. make pcr run PCR regression script and save result
8. make plsr run PLSR regression script and save result
9. make regressions run all regression model scripts together
10. make session run session info script and store system & package infromation into session-info.txt
11. make report assemble report from Rmd files in sections and transform to PDF format
12. make slides create slides from Rmd files
13. make clean remove old artifacts
14. make tests to run unit test in tests directory

Junyu Wang

Nichole Ann Rethmeier

ALl media content is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License

All code is licensed under MIT license.