Package MyMediaLite.ItemRecommendation

This namespace contains item recommenders and some helper classes for item recommendation. More...

Classes
class	BPRMF
	Matrix factorization model for item prediction (ranking) optimized for BPR More...
class	BPRSLIM
	Sparse Linear Methods (SLIM) for item prediction (ranking) optimized for BPR-Opt optimization criterion More...
class	Extensions
	Class that contains static methods for item prediction More...
class	ExternalItemRecommender
	Uses externally computed predictions More...
interface	IFoldInItemRecommender
	Item recommender that allows folding in new users More...
interface	IIncrementalItemRecommender
	Interface for item recommenders More...
class	IncrementalItemRecommender
	Base class for item recommenders that support incremental updates More...
class	ItemAttributeKNN
	k-nearest neighbor (kNN) item-based collaborative filtering using the correlation of the item attibutes More...
class	ItemKNN
	k-nearest neighbor (kNN) item-based collaborative filtering More...
class	ItemRecommender
	Abstract item recommender class that loads the (positive-only implicit feedback) training data into memory and provides flexible access to it. More...
interface	ITransductiveItemRecommender
	Interface for item recommenders that take into account some test data for training More...
class	KNN
	Base class for item recommenders that use some kind of k-nearest neighbors (kNN) model More...
class	LeastSquareSLIM
	Sparse Linear Methods (SLIM) for item prediction (ranking) optimized for the elastic net loss More...
class	MF
	Abstract class for matrix factorization based item predictors More...
class	MostPopular
	Most-popular item recommender More...
class	MostPopularByAttributes
	Recommend most popular items by attribute More...
class	MultiCoreBPRMF
	Matrix factorization for BPR on multiple cores More...
class	Random
	Random item recommender for use as experimental baseline More...
class	SLIM
	Abstract class for SLIM based item predictors proposed by Ning and Karypis More...
class	SoftMarginRankingMF
	Matrix factorization model for item prediction optimized for a soft margin (hinge) ranking loss, using stochastic gradient descent (as in BPR-MF). More...
class	UserAttributeKNN
	k-nearest neighbor (kNN) user-based collaborative filtering using the correlation of the user attibutes More...
class	UserKNN
	k-nearest neighbor user-based collaborative filtering More...
class	WeightedBPRMF
	Weigthed BPR-MF with frequency-adjusted sampling More...
class	WRMF
	Weighted matrix factorization method proposed by Hu et al. and Pan et al. More...
class	Zero
	Constant item recommender for use as experimental baseline. Always predicts a score of zero More...

Detailed Description

This namespace contains item recommenders and some helper classes for item recommendation.