Getting Started using IPython Notebooks

This section contains documentation generated from IPython notebooks that discuss different aspects of SMURFF.

• A first example running SMURFF
  • Downloading the data files
  • Having a look at the data
  • Running SMURFF
  • Plotting predictions versus actual values
• Input to SMURFF
  • Dense Train Input
  • Sparse Matrix Input
  • Tensor input
• Trying different Matrix Factorzation Methods
  • Downloading the files
  • Matrix Factorization without Side Information (BPMF)
  • Matrix Factorization with Side Information (Macau)
    • Macau univariate sampler
• Different noise models
  • Prepare train, test and side-info
  • Fixed noise
  • Adaptive noise
  • Binary matrices
    • Binary matrices with Side Info
• Inference with SMURFF
  • Saving models
  • Saved Model
  • Making predictions from a TrainSession
    • Predict all elements
    • Predict element in a sparse matrix
    • Predict just one element
    • Make predictions using side information
  • Making predictions from saved run
• Centering