search

mzakariaCERN / FunWithR

Some nice code scripts in R depectinve various ML models
0 stars 0 forks source link
decision-trees knn-classification naive-bayes-classifier r

readme

FunWithR

Some code scripts I have been working on using R:

  1. 1_kNN
    Folder contains some exploritary work using kNN classification for Cancer data link

  2. 2_NP
    Folder contains a study on using Naive Bayed for SMS classification link

  3. 3_DT
    Folder contains a study on using Decision Trees for default prediction on data from a bank link

  4. 4_RL Folder contains a study using Rule LEarners (R1 and JRip) to classify mushrooms link

  5. 5_Regression Folder contains a study using Linear Regression to predict medical expenses link

  6. 6_CART_ModelTrees using CART and model trees to estimate quality of wine link

  7. 7_

All studies use caret library for tuning using ROC

Additional Resource:

  1. UsefullLibraries: lists the libraries that I have encountered and deemed useful. Sort of a bookmark link

  2. Scrapping_NHL.R: Designed to scrap data from NHL website for minutes played by each player of the team for that game. Please see comments in the source file.

  3. CalculateNHLBands.Rmd: A short example on how to calculate the acceleration bands from a 5-column csv file. After finding the average acceleration, we calcualte the Standard deviation, and then make thresholds (band) using 0.5, 1.0, and 1.5 times the standard deviation.