Package 'mlsauce'

Title: Miscellaneous Statistical/Machine Learning stuff
Description: Miscellaneous Statistical/Machine Learning stuff.
Authors: T. Moudiki
Maintainer: T. Moudiki <[email protected]>
License: BSD_3_clause Clear + file LICENSE
Version: 0.37.2
Built: 2026-06-02 08:09:18 UTC
Source: https://github.com/Techtonique/mlsauce_r

Help Index

AdaOpt classifier

Description

AdaOpt classifier

Usage

AdaOpt(
  n_iterations = 50L,
  learning_rate = 0.3,
  reg_lambda = 0.1,
  reg_alpha = 0.5,
  eta = 0.01,
  gamma = 0.01,
  k = 3L,
  tolerance = 0,
  n_clusters = 0,
  batch_size = 100L,
  row_sample = 1,
  type_dist = "euclidean-f",
  cache = TRUE,
  n_clusters_input = 0,
  clustering_method = "kmeans",
  cluster_scaling = "standard",
  seed = 123L,
  venv_path = "./venv",
  ...
)

Arguments

n_iterations

number of iterations of the optimizer at training time
learning_rate

controls the speed of the optimizer at training time
reg_lambda

L2 regularization parameter for successive errors in the optimizer (at training time)
reg_alpha

L1 regularization parameter for successive errors in the optimizer (at training time)
eta

controls the slope in gradient descent (at training time)
gamma

controls the step size in gradient descent (at training time)
k

number of nearest neighbors selected at test time for classification
tolerance

controls early stopping in gradient descent (at training time)
n_clusters

number of clusters, if MiniBatch k-means is used at test time (for faster prediction)
batch_size

size of the batch, if MiniBatch k-means is used at test time (for faster prediction)
row_sample

percentage of rows chosen from training set (by stratified subsampling, for faster prediction)
type_dist

distance used for finding the nearest neighbors; currently euclidean-f (euclidean distances calculated as whole), euclidean (euclidean distances calculated row by row), cosine (cosine distance)
cache

if the nearest neighbors are cached or not, for faster retrieval in subsequent calls
n_clusters_input

number of clusters a priori on inpu data
clustering_method

either "kmeans" or "gmm" (Gaussian mixture)
cluster_scaling

either 'standard', 'minmax', 'robust'
seed

reproducibility seed for initial weak learner and clustering

Value

An object of class AdaOpt

Examples

## Not run: 
library(datasets)

X <- as.matrix(iris[, 1:4])
y <- as.integer(iris[, 5]) - 1L

n <- dim(X)[1]
p <- dim(X)[2]
set.seed(21341)
train_index <- sample(x = 1:n, size = floor(0.8*n), replace = TRUE)
test_index <- -train_index
X_train <- as.matrix(iris[train_index, 1:4])
y_train <- as.integer(iris[train_index, 5]) - 1L
X_test <- as.matrix(iris[test_index, 1:4])
y_test <- as.integer(iris[test_index, 5]) - 1L


obj <- mlsauce::AdaOpt()

print(obj$get_params())

obj$fit(X_train, y_train)

print(obj$score(X_test, y_test))

## End(Not run)

GenericBoosting classifier

Description

GenericBoosting classifier

Usage

GenericBoostingClassifier(
  base_model = NULL,
  n_estimators = 100L,
  learning_rate = 0.1,
  n_hidden_features = 5L,
  row_sample = 1,
  col_sample = 1,
  dropout = 0,
  tolerance = 1e-04,
  direct_link = 1L,
  verbose = 1L,
  seed = 123L,
  activation = "relu",
  n_clusters = 0,
  clustering_method = "kmeans",
  cluster_scaling = "standard",
  degree = NULL,
  weights_distr = "uniform",
  venv_path = "./venv",
  ...
)

Arguments

base_model:

object, base model to be boosted.
n_estimators:

int, number of boosting iterations.
learning_rate:

float, controls the learning speed at training time.
n_hidden_features:

int
number

of nodes in successive hidden layers.
row_sample:

float, percentage of rows chosen from the training set.
col_sample:

float, percentage of columns chosen from the training set.
dropout:

float, percentage of nodes dropped from the training set.
tolerance:

float, controls early stopping in gradient descent (at training time).
direct_link:

bool, indicates whether the original features are included (True) in model's fitting or not (False).
verbose:

int, progress bar (yes = 1) or not (no = 0) (currently).
seed:

int, reproducibility seed for nodes_sim=='uniform', clustering and dropout.
activation:

str, activation function: currently 'relu', 'relu6', 'sigmoid', 'tanh'
n_clusters:

int, number of clusters for clustering.
clustering_method:

str, clustering method: currently 'kmeans', 'gmm' (Gaussian Mixture Model)
cluster_scaling:

str, scaling method for clustering: currently 'standard', 'minmax', 'robust'
degree:

int, degree of polynomial interactions features.
weights_distr:

str, distribution of weights for the hidden layer currently 'uniform', 'gaussian'

Value

An object of class GenericBoostingClassifier

Examples

# See https://thierrymoudiki.github.io/blog/2024/10/14/r/genericboosting-r for advanced examples
## Not run: 
library(datasets)

X <- as.matrix(iris[, 1:4])
y <- as.integer(iris[, 5]) - 1L

n <- dim(X)[1]
p <- dim(X)[2]
set.seed(21341)
train_index <- sample(x = 1:n, size = floor(0.8*n), replace = TRUE)
test_index <- -train_index
X_train <- as.matrix(X[train_index, ])
y_train <- as.integer(y[train_index])
X_test <- as.matrix(X[test_index, ])
y_test <- as.integer(y[test_index])

sklearn <- nnetsauce::get_sklearn()

obj2 <- sklearn$tree$DecisionTreeRegressor()

obj <- mlsauce::GenericBoostingClassifier(obj2)

print(obj$get_params())

obj$fit(X_train, y_train)

print(obj$score(X_test, y_test))

sklearn <- nnetsauce::get_sklearn()

obj2 <- sklearn$linear_model$LinearRegression()

obj <- mlsauce::GenericBoostingClassifier(obj2)

print(obj$get_params())

obj$fit(X_train, y_train)

print(obj$score(X_test, y_test))

## End(Not run)

GenericBoosting Regressor

Description

GenericBoosting Regressor

Usage

GenericBoostingRegressor(
  base_model = NULL,
  n_estimators = 100L,
  learning_rate = 0.1,
  n_hidden_features = 5L,
  row_sample = 1,
  col_sample = 1,
  dropout = 0,
  tolerance = 1e-04,
  direct_link = 1L,
  verbose = 1L,
  seed = 123L,
  activation = "relu",
  n_clusters = 0,
  clustering_method = "kmeans",
  cluster_scaling = "standard",
  degree = NULL,
  weights_distr = "uniform",
  venv_path = "./venv",
  ...
)

Arguments

base_model:

object, base model to be boosted.
n_estimators:

int, number of boosting iterations.
learning_rate:

float, controls the learning speed at training time.
n_hidden_features:

int
number

of nodes in successive hidden layers.
row_sample:

float, percentage of rows chosen from the training set.
col_sample:

float, percentage of columns chosen from the training set.
dropout:

float, percentage of nodes dropped from the training set.
tolerance:

float, controls early stopping in gradient descent (at training time).
direct_link:

bool, indicates whether the original features are included (True) in model's fitting or not (False).
verbose:

int, progress bar (yes = 1) or not (no = 0) (currently).
seed:

int, reproducibility seed for nodes_sim=='uniform', clustering and dropout.
activation:

str, activation function: currently 'relu', 'relu6', 'sigmoid', 'tanh'
n_clusters:

int, number of clusters for clustering.
clustering_method:

str, clustering method: currently 'kmeans', 'gmm' (Gaussian Mixture Model)
cluster_scaling:

str, scaling method for clustering: currently 'standard', 'minmax', 'robust'
degree:

int, degree of polynomial interactions features.
weights_distr:

str, distribution of weights for the hidden layer currently 'uniform', 'gaussian'

Value

An object of class GenericBoostingRegressor

Examples

# See https://thierrymoudiki.github.io/blog/2024/10/14/r/genericboosting-r for advanced examples

## Not run: 
library(datasets)

X <- as.matrix(datasets::mtcars[, -1])
y <- as.integer(datasets::mtcars[, 1])

n <- dim(X)[1]
p <- dim(X)[2]
set.seed(21341)
train_index <- sample(x = 1:n, size = floor(0.8*n), replace = TRUE)
test_index <- -train_index
X_train <- as.matrix(X[train_index, ])
y_train <- as.double(y[train_index])
X_test <- as.matrix(X[test_index, ])
y_test <- as.double(y[test_index])


sklearn <- nnetsauce::get_sklearn()

obj2 <- sklearn$linear_model$LinearRegression()

obj <- mlsauce::GenericBoostingRegressor(obj2)

print(obj$get_params())

obj$fit(X_train, y_train)

print(sqrt(mean((obj$predict(X_test) - y_test)**2)))

## End(Not run)

Lasso regressor

Description

Lasso regressor

Usage

LassoRegressor(
  reg_lambda = 0.1,
  max_iter = 10L,
  tol = 0.001,
  venv_path = "./venv",
  ...
)

Arguments

reg_lambda

L1 regularization parameter
max_iter

number of iterations of lasso shooting algorithm.
tol

tolerance for convergence of lasso shooting algorithm.

Value

An object of class Lasso

Examples

## Not run: 
library(datasets)

X <- as.matrix(datasets::mtcars[, -1])
y <- as.integer(datasets::mtcars[, 1])

n <- dim(X)[1]
p <- dim(X)[2]
set.seed(21341)
train_index <- sample(x = 1:n, size = floor(0.8*n), replace = TRUE)
test_index <- -train_index
X_train <- as.matrix(X[train_index, ])
y_train <- as.double(y[train_index])
X_test <- as.matrix(X[test_index, ])
y_test <- as.double(y[test_index])

obj <- mlsauce::LassoRegressor()

print(obj$get_params())

obj$fit(X_train, y_train)

print(obj$score(X_test, y_test))

## End(Not run)

LazyBoostingClassifier

Description

Lazy Generic Boosting Classifier (AutoML Hold-out set validation)

Usage

LazyBoostingClassifier(
  verbose = 0,
  ignore_warnings = TRUE,
  custom_metric = NULL,
  predictions = FALSE,
  sort_by = "Accuracy",
  random_state = 42L,
  estimators = "all",
  preprocess = FALSE,
  n_jobs = NULL,
  venv_path = "./venv",
  ...
)

Arguments

verbose:

int, progress bar (yes = 1) or not (no = 0) (currently).
ignore_warnings:

bool, ignore warnings.
custom_metric:

function, custom metric.
predictions:

bool, return predictions.
sort_by:

str, sort by metric.
random_state:

int, random state.
estimators:

str, estimators to use. List of names for custom, or just 'all'.
preprocess:

bool, preprocess data or not.
n_jobs:

int, number of jobs.

Value

LazyBoostingClassifier object

Examples

# See https://thierrymoudiki.github.io/blog/2024/10/14/r/genericboosting-r for advanced examples

## Not run:  
library(mlsauce)
library(datasets)

data(iris)

X <- as.matrix(iris[, 1:4])
y <- as.integer(iris[, 5]) - 1L

n <- dim(X)[1]
p <- dim(X)[2]

set.seed(21341)
train_index <- sample(x = 1:n, size = floor(0.8*n), replace = TRUE)
test_index <- -train_index

X_train <- as.matrix(X[train_index, ])
y_train <- as.integer(y[train_index])
X_test <- as.matrix(X[test_index, ])
y_test <- as.integer(y[test_index])

obj <- LazyBoostingClassifier(verbose=0, ignore_warnings=TRUE,
                              custom_metric=NULL, preprocess=FALSE)

obj$fit(X_train, X_test, y_train, y_test)

## End(Not run)

LazyBoostingRegressor

Description

Lazy Generic Boosting Regressor (AutoML Hold-out set validation)

Usage

LazyBoostingRegressor(
  verbose = 0,
  ignore_warnings = TRUE,
  custom_metric = NULL,
  predictions = FALSE,
  sort_by = "RMSE",
  random_state = 42L,
  estimators = "all",
  preprocess = FALSE,
  n_jobs = NULL,
  venv_path = "./venv",
  ...
)

Arguments

verbose:

int, progress bar (yes = 1) or not (no = 0) (currently).
ignore_warnings:

bool, ignore warnings.
custom_metric:

function, custom metric.
predictions:

bool, return predictions.
sort_by:

str, sort by metric.
random_state:

int, random state.
estimators:

str, estimators to use. List of names for custom, or just 'all'.
preprocess:

bool, preprocess data or not.
n_jobs:

int, number of jobs.

Value

LazyBoostingRegressor object

Examples

# See https://thierrymoudiki.github.io/blog/2024/10/14/r/genericboosting-r for advanced examples
## Not run: 
library(mlsauce)
library(datasets)

data(mtcars)

X <- as.matrix(mtcars[, -1])
y <- as.integer(mtcars[, 1]) 

n <- dim(X)[1]
p <- dim(X)[2]

set.seed(21341)
train_index <- sample(x = 1:n, size = floor(0.8*n), replace = TRUE)
test_index <- -train_index

X_train <- as.matrix(X[train_index, ])
y_train <- as.integer(y[train_index])
X_test <- as.matrix(X[test_index, ])
y_test <- as.integer(y[test_index])

obj <- LazyBoostingRegressor(verbose=0, ignore_warnings=TRUE,
                              custom_metric=NULL, preprocess=FALSE)

obj$fit(X_train, X_test, y_train, y_test)

## End(Not run)

LSBoost classifier

Description

LSBoost classifier

Usage

LSBoostClassifier(
  n_estimators = 100L,
  learning_rate = 0.1,
  n_hidden_features = 5L,
  reg_lambda = 0.1,
  row_sample = 1,
  col_sample = 1,
  dropout = 0,
  tolerance = 1e-04,
  direct_link = 1L,
  verbose = 1L,
  seed = 123L,
  solver = c("ridge", "lasso"),
  activation = "relu",
  n_clusters = 0L,
  clustering_method = "kmeans",
  cluster_scaling = "standard",
  degree = 1L,
  weights_distr = "uniform",
  venv_path = "./venv",
  ...
)

Arguments

n_estimators:

int, number of boosting iterations.
learning_rate:

float, controls the learning speed at training time.
n_hidden_features:

int
number

of nodes in successive hidden layers.
reg_lambda:

float, L2 regularization parameter for successive errors in the optimizer (at training time).
row_sample:

float, percentage of rows chosen from the training set.
col_sample:

float, percentage of columns chosen from the training set.
dropout:

float, percentage of nodes dropped from the training set.
tolerance:

float, controls early stopping in gradient descent (at training time).
direct_link:

bool, indicates whether the original features are included (True) in model's fitting or not (False).
verbose:

int, progress bar (yes = 1) or not (no = 0) (currently).
seed:

int, reproducibility seed for nodes_sim=='uniform', clustering and dropout.
solver:

str, type of 'weak' learner; currently in ('ridge', 'lasso')
activation:

str, activation function: currently 'relu', 'relu6', 'sigmoid', 'tanh'
n_clusters:

int, number of clusters for clustering.
clustering_method:

str, clustering method: currently 'kmeans', 'gmm' (Gaussian Mixture Model)
cluster_scaling:

str, scaling method for clustering: currently 'standard', 'minmax', 'robust'
degree:

int, degree of polynomial interactions features.
weights_distr:

str, distribution of weights for the hidden layer currently 'uniform', 'gaussian'

Value

An object of class LSBoostClassifier

Examples

library(datasets)

X <- as.matrix(iris[, 1:4])
y <- as.integer(iris[, 5]) - 1L

n <- dim(X)[1]
p <- dim(X)[2]
set.seed(21341)
train_index <- sample(x = 1:n, size = floor(0.8*n), replace = TRUE)
test_index <- -train_index
X_train <- as.matrix(X[train_index, ])
y_train <- floor(y[train_index])
X_test <- as.matrix(X[test_index, ])
y_test <- floor(y[test_index])

## Not run: 
obj <- mlsauce::LSBoostClassifier()

print(obj$get_params())

obj$fit(X_train, y_train)

print(obj$score(X_test, y_test))

## End(Not run)

LSBoost Regressor

Description

LSBoost Regressor

Usage

LSBoostRegressor(
  n_estimators = 100L,
  learning_rate = 0.1,
  n_hidden_features = 5L,
  reg_lambda = 0.1,
  row_sample = 1,
  col_sample = 1,
  dropout = 0,
  tolerance = 1e-04,
  direct_link = 1L,
  verbose = 1L,
  seed = 123L,
  solver = c("ridge", "lasso"),
  activation = "relu",
  n_clusters = 0L,
  clustering_method = "kmeans",
  cluster_scaling = "standard",
  degree = 1L,
  weights_distr = "uniform",
  venv_path = "./venv",
  ...
)

Arguments

n_estimators:

int, number of boosting iterations.
learning_rate:

float, controls the learning speed at training time.
n_hidden_features:

int
number

of nodes in successive hidden layers.
reg_lambda:

float, L2 regularization parameter for successive errors in the optimizer (at training time).
row_sample:

float, percentage of rows chosen from the training set.
col_sample:

float, percentage of columns chosen from the training set.
dropout:

float, percentage of nodes dropped from the training set.
tolerance:

float, controls early stopping in gradient descent (at training time).
direct_link:

bool, indicates whether the original features are included (True) in model's fitting or not (False).
verbose:

int, progress bar (yes = 1) or not (no = 0) (currently).
seed:

int, reproducibility seed for nodes_sim=='uniform', clustering and dropout.
solver:

str, type of 'weak' learner; currently in ('ridge', 'lasso')
activation:

str, activation function: currently 'relu', 'relu6', 'sigmoid', 'tanh'
n_clusters:

int, number of clusters for clustering.
clustering_method:

str, clustering method: currently 'kmeans', 'gmm' (Gaussian Mixture Model)
cluster_scaling:

str, scaling method for clustering: currently 'standard', 'minmax', 'robust'
degree:

int, degree of polynomial interactions features.
weights_distr:

str, distribution of weights for the hidden layer currently 'uniform', 'gaussian'

Value

An object of class LSBoostRegressor

Examples

## Not run: 
library(datasets)

X <- as.matrix(datasets::mtcars[, -1])
y <- as.integer(datasets::mtcars[, 1])

n <- dim(X)[1]
p <- dim(X)[2]
set.seed(21341)
train_index <- sample(x = 1:n, size = floor(0.8*n), replace = TRUE)
test_index <- -train_index
X_train <- as.matrix(X[train_index, ])
y_train <- as.double(y[train_index])
X_test <- as.matrix(X[test_index, ])
y_test <- as.double(y[test_index])

obj <- mlsauce::LSBoostRegressor()

print(obj$get_params())

obj$fit(X_train, y_train)

print(obj$score(X_test, y_test))

## End(Not run)

Ridge regressor

Description

Ridge regressor

Usage

RidgeRegressor(reg_lambda = 0.1, venv_path = "./venv", ...)

Arguments

reg_lambda

L2 regularization parameter

Value

An object of class Ridge

Examples

## Not run: 
library(datasets)

X <- as.matrix(datasets::mtcars[, -1])
y <- as.integer(datasets::mtcars[, 1])

n <- dim(X)[1]
p <- dim(X)[2]
set.seed(21341)
train_index <- sample(x = 1:n, size = floor(0.8*n), replace = TRUE)
test_index <- -train_index
X_train <- as.matrix(X[train_index, ])
y_train <- as.double(y[train_index])
X_test <- as.matrix(X[test_index, ])
y_test <- as.double(y[test_index])


obj <- mlsauce::RidgeRegressor()

print(obj$get_params())

obj$fit(X_train, y_train)

print(obj$score(X_test, y_test))

## End(Not run)

Stump classifier

Description

Stump classifier

Usage

StumpClassifier(bins = "auto", venv_path = "./venv", ...)

Arguments

bins:

int, number of histogram bins.

Value

An object of class StumpClassifier

Examples

## Not run: 
library(datasets)

X <- as.matrix(iris[, 1:4])
y <- as.integer(iris[, 5]) - 1L

n <- dim(X)[1]
p <- dim(X)[2]
set.seed(21341)
train_index <- sample(x = 1:n, size = floor(0.8*n), replace = TRUE)
test_index <- -train_index
X_train <- as.matrix(iris[train_index, 1:4])
y_train <- as.integer(iris[train_index, 5]) - 1L
X_test <- as.matrix(iris[test_index, 1:4])
y_test <- as.integer(iris[test_index, 5]) - 1L


obj <- mlsauce::StumpClassifier()

print(obj$get_params())

obj$fit(X_train, y_train)

print(obj$score(X_test, y_test))

## End(Not run)