Provides implementations of several popular recommendation systems. They can process standard recommendation datasets (user/item matrix) as input and generate rating predictions and recommendation lists. Standard algorithm implementations included in this package are: Global/Item/User-Average baselines, Item-Based KNN, FunkSVD, BPR and weighted ALS. They can be assessed according to the standard offline evaluation methodology for recommender systems using measures such as MAE, RMSE, Precision, Recall, AUC, NDCG, RankScore and coverage measures. The package is intended for rapid prototyping of recommendation algorithms and education purposes.
|Author||Ludovik Çoba [aut, cre, cph], Markus Zanker [ctb]|
|Date of publication||2016-06-27 17:31:21|
|Maintainer||Ludovik Çoba <firstname.lastname@example.org>|
algAverageClass: Baseline algorithms exploiting global/item and user averages.
BPRclass: Bayesian Personalized Ranking based model.
dataSet-class: Dataset class.
defineData: Define dataset.
evalModel: Creating the evaluation model.
evalModel-class: Evaluation model.
evalPred: Evaluates the requested prediction algorithm.
evalrec: Evaluates the requested recommendation algorithm.
getAUC: Returns the Area under the ROC curve.
IBclass: Item based model.
mlLatest100k: Movielens Latest
nDCG: Normalized Discounted Cumulative Gain
PPLclass: Popularity based model.
predict: Generate predictions.
rankScore: Rank Score
recommend: Generate recommendation.
recResultsClass: Results of a recommendation.
rrecsys: Create a recommender system.
setStoppingCriteria: Set stopping criteria.
SVDclass: SVD model.
wALSclass: Weighted Alternating Least Squares based model.