This is the library that I use to collect common machine learning algorithm I use for research project and industrial practice.
The installation of this package is simple. We recommend to use devtools to install from Github.
# Note: you must have these packages: # devtools,quantmod,stats,xts,TTR,knitr,gtools,keras,naivebayes,randomForest,iRF,BayesTree,gbm,pROC # Please install them if you do not have them. # Install Package: Yin's Library (i.e. YinsLibrary) devtools::install_github("yiqiao-yin/YinsLibrary")
This package has all functions collected in R folder. Almost all functions following the following format:
- input $X$: this is the explanatory variables in the data set;
- input $y$: this is the response variable in the data set;
- input cutoff: this is a numerical value from 0 to 1 (default value is 0.9), implying that the algorithm will take the first 90% of the observation as training and the rest as testing;
- input parameters: this is dependent on each function (ex: for trees, there are number of trees, for SVM, there is gamma, etc..);
- input cutoff.coefficient: this is a value with default at 1, which means the algorithm is setting exactly mean of predicted scores as cutoff to convert scores into binary values.
Yiqiao Yin has been a Research Assistant at Columbia University since 2017. Prior to his current position, he has been a Researcher at Simon Business School with professors from AQR Capital.