Resampling, Selection, & Regularization

(ns index)

1 Assignment 3: Resampling, Selection, & Regularization

Instructions:

• Select a dataset suitable to regression at your discretion
• Using built in functionality, apply the following algorithms to your data using 5 fold cross validation and RMSE:
      Lasso Regression
      Ridge Regression
      * Linear or SGD Regression w/ KBestFeature Selection
  All final code should be in the main.py file. Modify the “README.md” file to include the following sections:
      * Summary: Recap what was learned in class. This should be thorough enough that someone else could understand what the class was about
  Questions (Answer these):
  What were the best hyperparameters for Lasso/Ridge Regression?
  Which features did Lasso/Ridge Regression choose?
  What was the optimal K in K Best Features Selection?
  Make sure that your README file is formatted properly and is visually appealing. It should be free of grammatical errors, punctuation errors, capitalization issues, etc. Sentences should be complete.

What I did:

• Selected Boston Housing Prices. One of my favorite regression datsets to play around with and teach.
• Looked at the data with different transformations in mind. Standardizing and Tukey’s ladder transformation.
• Implemented regularization techniques–ridge and lasso–and created a best subset model of the 13 regressors.
• Compared fitting model to raw vs fitting model to Tukey tranformed data.

source: src/index.clj