I am involved in a data science project with an online betting company (hence the gambleR project name). I thought I would share the R code I have been writing for prediction models. There is of course no specifics about the data and you can easily reuse all the code.
Bunch of functions to compute the precision, recall and f-measure of a vector of predictions compared to the actuals values. Can be restricted to a given class.
Using the ROCR package, it can also compute successive precision/recall values for different cutoffs.
Examples of how to use trees (CART, C4.5, J4.8, conditional inference), random forests, SVM and Naive Bayes to train classification models.
Examples of how to apply trees (CART, C4.5, J4.8, conditional inference), random forests, SVM and Naive Bayes to predict class or probabilities.