Class: Rumale::PolynomialModel::FactorizationMachineRegressor
- Inherits:
-
BaseFactorizationMachine
- Object
- BaseFactorizationMachine
- Rumale::PolynomialModel::FactorizationMachineRegressor
- Includes:
- Base::Regressor
- Defined in:
- lib/rumale/polynomial_model/factorization_machine_regressor.rb
Overview
FactorizationMachineRegressor is a class that implements Factorization Machine with stochastic gradient descent (SGD) optimization.
Reference
-
Rendle, “Factorization Machines with libFM,” ACM TIST, vol. 3 (3), pp. 57:1–57:22, 2012.
-
-
Rendle, “Factorization Machines,” Proc. ICDM’10, pp. 995–1000, 2010.
-
Instance Attribute Summary collapse
-
#bias_term ⇒ Numo::DFloat
readonly
Return the bias term for Factoriazation Machine.
-
#factor_mat ⇒ Numo::DFloat
readonly
Return the factor matrix for Factorization Machine.
-
#rng ⇒ Random
readonly
Return the random generator for random sampling.
-
#weight_vec ⇒ Numo::DFloat
readonly
Return the weight vector for Factorization Machine.
Attributes included from Base::BaseEstimator
Instance Method Summary collapse
-
#fit(x, y) ⇒ FactorizationMachineRegressor
Fit the model with given training data.
-
#initialize(n_factors: 2, reg_param_linear: 1.0, reg_param_factor: 1.0, max_iter: 1000, batch_size: 10, optimizer: nil, n_jobs: nil, random_seed: nil) ⇒ FactorizationMachineRegressor
constructor
Create a new regressor with Factorization Machine.
-
#marshal_dump ⇒ Hash
Dump marshal data.
-
#marshal_load(obj) ⇒ nil
Load marshal data.
-
#predict(x) ⇒ Numo::DFloat
Predict values for samples.
Methods included from Base::Regressor
Constructor Details
#initialize(n_factors: 2, reg_param_linear: 1.0, reg_param_factor: 1.0, max_iter: 1000, batch_size: 10, optimizer: nil, n_jobs: nil, random_seed: nil) ⇒ FactorizationMachineRegressor
Create a new regressor with Factorization Machine.
55 56 57 58 59 60 61 62 63 64 |
# File 'lib/rumale/polynomial_model/factorization_machine_regressor.rb', line 55 def initialize(n_factors: 2, reg_param_linear: 1.0, reg_param_factor: 1.0, max_iter: 1000, batch_size: 10, optimizer: nil, n_jobs: nil, random_seed: nil) check_params_float(reg_param_linear: reg_param_linear, reg_param_factor: reg_param_factor) check_params_integer(n_factors: n_factors, max_iter: max_iter, batch_size: batch_size) check_params_type_or_nil(Integer, n_jobs: n_jobs, random_seed: random_seed) check_params_positive(n_factors: n_factors, reg_param_linear: reg_param_linear, reg_param_factor: reg_param_factor, max_iter: max_iter, batch_size: batch_size) keywd_args = method(:initialize).parameters.map { |_t, arg| [arg, binding.local_variable_get(arg)] }.to_h.merge(loss: nil) super(keywd_args) end |
Instance Attribute Details
#bias_term ⇒ Numo::DFloat (readonly)
Return the bias term for Factoriazation Machine.
35 36 37 |
# File 'lib/rumale/polynomial_model/factorization_machine_regressor.rb', line 35 def bias_term @bias_term end |
#factor_mat ⇒ Numo::DFloat (readonly)
Return the factor matrix for Factorization Machine.
27 28 29 |
# File 'lib/rumale/polynomial_model/factorization_machine_regressor.rb', line 27 def factor_mat @factor_mat end |
#rng ⇒ Random (readonly)
Return the random generator for random sampling.
39 40 41 |
# File 'lib/rumale/polynomial_model/factorization_machine_regressor.rb', line 39 def rng @rng end |
#weight_vec ⇒ Numo::DFloat (readonly)
Return the weight vector for Factorization Machine.
31 32 33 |
# File 'lib/rumale/polynomial_model/factorization_machine_regressor.rb', line 31 def weight_vec @weight_vec end |
Instance Method Details
#fit(x, y) ⇒ FactorizationMachineRegressor
Fit the model with given training data.
71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 |
# File 'lib/rumale/polynomial_model/factorization_machine_regressor.rb', line 71 def fit(x, y) check_sample_array(x) check_tvalue_array(y) check_sample_tvalue_size(x, y) n_outputs = y.shape[1].nil? ? 1 : y.shape[1] _n_samples, n_features = x.shape if n_outputs > 1 @factor_mat = Numo::DFloat.zeros(n_outputs, @params[:n_factors], n_features) @weight_vec = Numo::DFloat.zeros(n_outputs, n_features) @bias_term = Numo::DFloat.zeros(n_outputs) if enable_parallel? models = parallel_map(n_outputs) { |n| partial_fit(x, y[true, n]) } n_outputs.times { |n| @factor_mat[n, true, true], @weight_vec[n, true], @bias_term[n] = models[n] } else n_outputs.times { |n| @factor_mat[n, true, true], @weight_vec[n, true], @bias_term[n] = partial_fit(x, y[true, n]) } end else @factor_mat, @weight_vec, @bias_term = partial_fit(x, y) end self end |
#marshal_dump ⇒ Hash
Dump marshal data.
113 114 115 116 117 118 119 |
# File 'lib/rumale/polynomial_model/factorization_machine_regressor.rb', line 113 def marshal_dump { params: @params, factor_mat: @factor_mat, weight_vec: @weight_vec, bias_term: @bias_term, rng: @rng } end |
#marshal_load(obj) ⇒ nil
Load marshal data.
123 124 125 126 127 128 129 130 |
# File 'lib/rumale/polynomial_model/factorization_machine_regressor.rb', line 123 def marshal_load(obj) @params = obj[:params] @factor_mat = obj[:factor_mat] @weight_vec = obj[:weight_vec] @bias_term = obj[:bias_term] @rng = obj[:rng] nil end |
#predict(x) ⇒ Numo::DFloat
Predict values for samples.
100 101 102 103 104 105 106 107 108 109 |
# File 'lib/rumale/polynomial_model/factorization_machine_regressor.rb', line 100 def predict(x) check_sample_array(x) linear_term = @bias_term + x.dot(@weight_vec.transpose) factor_term = if @weight_vec.shape[1].nil? 0.5 * (@factor_mat.dot(x.transpose)**2 - (@factor_mat**2).dot(x.transpose**2)).sum(0) else 0.5 * (@factor_mat.dot(x.transpose)**2 - (@factor_mat**2).dot(x.transpose**2)).sum(1).transpose end linear_term + factor_term end |