Abstract class for matrix factorization based item predictors
Declaring type: MF
Namespace: MyMediaLite.ItemRecommendation
Assembly: MyMediaLite
Collapse/Expand Protected Fields
  Name Description
Protected Field init_mean Mean of the normal distribution used to initialize the latent factors
Protected Field init_stdev Standard deviation of the normal distribution used to initialize the latent factors
Protected Field item_factors Latent item factor matrix
Protected Field num_factors Number of latent factors per user/item
Protected Field user_factors Latent user factor matrix
Top
Collapse/Expand Protected Constructors
  Name Description
Protected Method MF

There is no summary.
Top
Collapse/Expand Public Methods
  Name Description
Public Method Abstract ComputeFit Computes the fit (optimization criterion) on the training data
Public Method Abstract Iterate Iterate once over the data
Public Method Virtual LoadModel

There is no summary.
Public Method Virtual Predict Predict the weight for a given user-item combination.
Public Method Virtual SaveModel

There is no summary.
Public Method Virtual Train

There is no summary.
Top
Collapse/Expand Public Properties
  Name Description
Public Property InitMean Mean of the normal distribution used to initialize the latent factors
Public Property InitStdev Standard deviation of the normal distribution used to initialize the latent factors
Public Property NumFactors Number of latent factors per user/item
Public Property NumIter Number of iterations over the training data
Top