FactorizationMachine

Creates second order Factorization Machines model

datasets

Implements several algorithms for
supervised learning on sparse data and many
matrix factorizations of sparse matrices
(with a focus on applications for recommender systems).
All algorithms work on sparse matrices. Also they extensively
use BLAS and LAPACK and parallelized with OpenMP.
Implementations are reasonably fast and nicely work with large
datasets (millions of rows and millions of columns).
List of algorithms for supervised learning:
1) Elastic net regression via Follow The Proximally-Regularized leader algorithm
2) Second order Factorization Machines via stochastic gradient descent with
adaptive learning rates. Allows to learn model parameters out-of-core.
Fast - asynchronous parallel, SIMD accelerated.
List of algorithms for matrix factorization:
1) Weighted Regularazied Matrix Factorization with Alternating Least Squares (ALS)
for implicit feedback (inculding approximate Conjugate Gradient solver).
Optional non-negativity (NNMF, non-negative matrix factorization).
2) Regularazied Matrix Factorization with ALS for explicit feedback
Optional non-negativity (NNMF, non-negative matrix factorization).
3) Fast Trunceate SVD and Soft-SVD via ALS
4) Soft-Impute via fast ALS and solution in SVD form
5) LinearFlow method which learns item-item similarity matrix from the data
Clustering:
1) kmeans from Armadillo library which provides smart (similar to kmeans++)
cluster initializations.
Misc utils/methods:
1) multithreaded `%*%` and `tcrossprod()` for `<dgRMatrix, matrix>`
2) multithreaded `%*%` and `crossprod()` for `<matrix, dgCMatrix>`

Dmitriy Selivanov

rsparse

Statistical Learning on Sparse Matrices

Drew Schmidt

Han Zhang

FactorizationMachine function

Format

For usage details see Methods, Arguments and Examples sections.<pre>
fm = FM$new(learning_rate_w = 0.2, rank = 8, lambda_w = 0, lambda_v = 0, task = c("classification", "regression")
 intercept = TRUE, learning_rate_v = learning_rate_w)
fm$partial_fit(x, y, ...)
fm$predict(x, ...)
</pre>

Usage

<dl class="dl-horizontal">
 <dt><code>FM$new(learning_rate_w = 0.2, rank = 8, lambda_w = 1e-6, lambda_v = 1e-6,
 task = c("classification", "regression"), intercept = TRUE, learning_rate_v = learning_rate_w)</code></dt><dd>Constructor
 for FactorizationMachines model. For description of arguments see Arguments section.</dd>
 <dt><code>$partial_fit(x, y, ...)</code></dt><dd>fits/updates model given input matrix <code>x</code> and target vector <code>y</code>.
 <code>x</code> shape = (n_samples, n_features)</dd>
 <dt><code>$predict(x, ...)</code></dt><dd>predicts output <code>x</code></dd>
</dl>

Methods

<dl class="dl-horizontal">
 <dt>fm</dt><dd><code>FM</code> object</dd>
 <dt>x</dt><dd>Input sparse matrix - native format is <code>Matrix::RsparseMatrix</code>.
 If <code>x</code> is in different format, model will try to convert it to <code>RsparseMatrix</code>
 with <code>as(x, "RsparseMatrix")</code> call</dd>
 <dt>learning_rate_w</dt><dd>learning rate for linear weights in AdaGrad SGD</dd>
 <dt>learning_rate_v</dt><dd>learning rate for interactions in AdaGrad SGD</dd>
 <dt>rank</dt><dd>rank of the latent dimension in factorization</dd>
 <dt>lambda_w</dt><dd>regularization parameter for linear terms</dd>
 <dt>lambda_v</dt><dd>regularization parameter for interactions terms</dd>
 <dt>n_features</dt><dd>number of features in model (number of columns in expected model matrix)</dd>
 <dt>task</dt><dd><code>"regression"</code> or <code>"classification"</code></dd>
</dl>

FactorizationMachine: Creates FactorizationMachine model.

Description

Usage

Format

Usage

Methods

Arguments