Forecast Combinations (ForecastCombinations)
Aim: Supports the most frequently used methods to combine forecasts. Among others: Simple average, Ordinary Least Squares, Least Absolute Deviation, Constrained Least Squares, Variance-based, Best Individual model, Complete subset regressions and Information-theoretic (information criteria based).
Searching the best regression model using R (correlation filter, data scaling, best regression model, etc.) (RRegrs)
The current tool is a collection of regression tools from R that could be used to search the best regression models for any dataset. The initial use of the script is aimed at finding QSAR models for chemoinformatics / nanotoxicology. The full R script will contain: Loading dataset, Filter dataset, Scaling dataset, Feature selection, Regression models, Summary with top models, Statistics of the best model, etc. The script will be modular in order to create flexible APIs.
Truncated Multivariate Normal (TruncatedNormal)
A collection of functions to deal with the truncated univariate and multivariate normal distributions.