xLearn - the high performance machine learning library - for Ruby
Supports:
- Linear models
- Factorization machines
- Field-aware factorization machines
Add this line to your application’s Gemfile:
gem "xlearn"
Prep your data
x = [[1, 2], [3, 4], [5, 6], [7, 8]]
y = [1, 2, 3, 4]
Train a model
model = XLearn::Linear.new(task: "reg")
model.fit(x, y)
Use XLearn::FM
for factorization machines and XLearn::FFM
for field-aware factorization machines
Make predictions
model.predict(x)
Save the model to a file
model.save_model("model.bin")
Load the model from a file
model.load_model("model.bin")
Save a text version of the model
model.save_txt("model.txt")
Pass a validation set
model.fit(x_train, y_train, eval_set: [x_val, y_val])
Train online
model.partial_fit(x_train, y_train)
Get the bias term, linear term, and latent factors
model.bias_term
model.linear_term
model.latent_factors # fm and ffm only
Pass parameters - default values below
XLearn::FM.new(
task: "binary", # binary (classification), reg (regression)
metric: nil, # acc, prec, recall, f1, auc, mae, mape, rmse, rmsd
lr: 0.2, # learning rate
lambda: 0.00002, # lambda for l2 regularization
k: 4, # latent factors for fm and ffm
alpha: 0.3, # hyper parameter for ftrl
beta: 1.0, # hyper parameter for ftrl
lambda_1: 0.00001, # hyper parameter for ftrl
lambda_2: 0.00002, # hyper parameter for ftrl
epoch: 10, # number of epochs
fold: 3, # number of folds
opt: "adagrad", # sgd, adagrad, ftrl
block_size: 500, # block size for on-disk training in MB
early_stop: true, # use early stopping
stop_window: 2, # size of stop window for early stopping
sign: false, # convert predition output to 0 and 1
sigmoid: false, # convert predition output using sigmoid
seed: 1 # random seed to shuffle data set
)
Cross-validation
model.cv(x, y)
Specify the number of folds
model.cv(x, y, folds: 5)
Data can be an array of arrays
[[1, 2, 3], [4, 5, 6]]
Or a Numo array
Numo::NArray.cast([[1, 2, 3], [4, 5, 6]])
Or a Rover data frame
Rover.read_csv("houses.csv")
Or a Daru data frame
Daru::DataFrame.from_csv("houses.csv")
For large datasets, read data directly from files
model.fit("train.txt", eval_set: "validate.txt")
model.predict("test.txt")
model.cv("train.txt")
For linear models and factorization machines, use CSV:
label,value_1,value_2,...,value_n
Or the libsvm
format (better for sparse data):
label index_1:value_1 index_2:value_2 ... index_n:value_n
You can also use commas instead of spaces for separators
For field-aware factorization machines, use the libffm
format:
label field_1:index_1:value_1 field_2:index_2:value_2 ...
You can also use commas instead of spaces for separators
You can also write predictions directly to a file
model.predict("test.txt", out_path: "predictions.txt")
This library is modeled after xLearn’s Scikit-learn API.
View the changelog
Everyone is encouraged to help improve this project. Here are a few ways you can help:
- Report bugs
- Fix bugs and submit pull requests
- Write, clarify, or fix documentation
- Suggest or add new features
To get started with development:
git clone https://github.com/ankane/xlearn-ruby.git
cd xlearn-ruby
bundle install
bundle exec rake vendor:all
bundle exec rake test