Skip to content
master
Switch branches/tags
Code

Latest commit

 

Git stats

Files

Permalink
Failed to load latest commit information.
Type
Name
Latest commit message
Commit time
lib
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

LIBMF Ruby

LIBMF - large-scale sparse matrix factorization - for Ruby

Check out Disco for higher-level collaborative filtering

Build Status

Installation

Add this line to your application’s Gemfile:

gem 'libmf'

Getting Started

Prep your data in the format row_index, column_index, value

data = Libmf::Matrix.new
data.push(0, 0, 5.0)
data.push(0, 2, 3.5)
data.push(1, 1, 4.0)

Create a model

model = Libmf::Model.new
model.fit(data)

Make predictions

model.predict(row_index, column_index)

Get the latent factors (these approximate the training matrix)

model.p_factors
model.q_factors

Get the bias (average of all elements in the training matrix)

model.bias

Save the model to a file

model.save_model("model.txt")

Load the model from a file

model.load_model("model.txt")

Pass a validation set

model.fit(data, eval_set: eval_set)

Cross-Validation

Perform cross-validation

model.cv(data)

Specify the number of folds

model.cv(data, folds: 5)

Parameters

Pass parameters - default values below

Libmf::Model.new(
  loss: :real_l2,         # loss function
  factors: 8,             # number of latent factors
  threads: 12,            # number of threads used
  bins: 25,               # number of bins
  iterations: 20,         # number of iterations
  lambda_p1: 0,           # coefficient of L1-norm regularization on P
  lambda_p2: 0.1,         # coefficient of L2-norm regularization on P
  lambda_q1: 0,           # coefficient of L1-norm regularization on Q
  lambda_q2: 0.1,         # coefficient of L2-norm regularization on Q
  learning_rate: 0.1,     # learning rate
  alpha: 0.1,             # importance of negative entries
  c: 0.0001,              # desired value of negative entries
  nmf: false,             # perform non-negative MF (NMF)
  quiet: false            # no outputs to stdout
)

Loss Functions

For real-valued matrix factorization

  • :real_l2 - squared error (L2-norm)
  • :real_l1 - absolute error (L1-norm)
  • :real_kl - generalized KL-divergence

For binary matrix factorization

  • :binary_log - logarithmic error
  • :binary_l2 - squared hinge loss
  • :binary_l1 - hinge loss

For one-class matrix factorization

  • :one_class_row - row-oriented pair-wise logarithmic loss
  • :one_class_col - column-oriented pair-wise logarithmic loss
  • :one_class_l2 - squared error (L2-norm)

Performance

For performance, read data directly from files

model.fit("train.txt", eval_set: "validate.txt")
model.cv("train.txt")

Data should be in the format row_index column_index value:

0 0 5.0
0 2 3.5
1 1 4.0

Numo

Get latent factors as Numo arrays

model.p_factors(format: :numo)
model.q_factors(format: :numo)

Resources

History

View the changelog

Contributing

Everyone is encouraged to help improve this project. Here are a few ways you can help:

To get started with development:

git clone --recursive https://github.com/ankane/libmf-ruby.git
cd libmf-ruby
bundle install
bundle exec rake vendor:all
bundle exec rake test

About

Large-scale sparse matrix factorization for Ruby

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages