mlpack_linear_svm(1)

📦
version 4.8.0 · 5 versions
, in ubuntu stonking

✨ mlpack_linear_svm ✨

okay so this is the manual page for mlpack_linear_svm(1), and bestie, if you've ever wanted a linear SVM that goes brrr, this is literally it. sit down, hydrate, let's get into it 🫶

💖 NAME 💖

mlpack_linear_svm — linear svm is an l2-regularized support vector machine. no cap. 🌸

🌈 SYNOPSIS 🌈

mlpack_linear_svm [-d double] [-E int] [-m unknown] [-l unknown] [-r double] [-n int] [-N bool] [-c int] [-O string] [-s int] [-S bool] [-a double] [-T unknown] [-L unknown] [-e double] [-t unknown] [-V bool] [-M unknown] [-P unknown] [-p unknown] [-h -v]

👑 DESCRIPTION 👑

so this is an implementation of linear SVMs that uses either L-BFGS or parallel SGD (stochastic gradient descent) to train the model. two iconic optimizers, one little program 💅

here's the whole vibe: this program lets you load a linear SVM model (via the --input_model_file (-m) parameter), train a linear SVM model given training data (specified with the --training_file (-t) parameter), or honestly do both of those things at once, because why choose. on top of that, it lets you do classification on a test dataset (specified with the --test_file (-T) parameter), and the classification results may be saved with the --predictions_file (-P) output parameter. and the trained linear SVM model itself? saved with the --output_model_file (-M) output parameter. slay ✨

the training data, if you specify it, may have class labels as its last dimension. alternately, the --labels_file (-l) parameter may be used to specify a separate vector of labels. your call, both are valid 🎀

when a model is being trained, there are so many options, and we're listing all of them because we keep every detail around here:

  • L2 regularization (which exists to prevent overfitting, the enemy) can be specified with the --lambda (-r) option.
  • the number of classes can be manually specified with the --num_classes (-c).
  • and if an intercept term is not desired in the model, the --no_intercept (-N) parameter can be specified.
  • margin of difference between correct class and other classes can be specified with the --delta (-d) option.
  • the optimizer used to train the model can be specified with the --optimizer (-O) parameter. available options are psgd (parallel stochastic gradient descent) and lbfgs (the L-BFGS optimizer).

there are also various parameters for the optimizer, and yes we're covering them:

  • the --max_iterations (-n) parameter specifies the maximum number of allowed iterations.
  • the --tolerance (-e) parameter specifies the tolerance for convergence.
  • for the parallel SGD optimizer, the --step_size (-a) parameter controls the step size taken at each iteration by the optimizer, and the maximum number of epochs is specified with --epochs (-E).
⚠️ heads up, and this is important: if the objective function for your data is oscillating between Inf and 0, the step size is probably too large. not a drill. 😭

there are more parameters for the optimizers, but the C++ interface must be used to access these. so like, iykyk, but that's a whole other door.

optionally, the model can be used to predict the labels for another matrix of data points, if --test_file (-T) is specified. and here's the cute part: the --test_file (-T) parameter can be specified without the --training_file (-t) parameter, so long as an existing linear SVM model is given with the --input_model_file (-m) parameter. the output predictions from the linear SVM model may be saved with the --predictions_file (-P) parameter. 🤝

🌟 DESCRIPTION: examples that actually eat 🌟

as an example, to train a linear SVM on the data data.csv with labels labels.csv with L2 regularization of 0.1, saving the model to lsvm_model.bin, the following command may be used:

$ mlpack_linear_svm --training_file data.csv --labels_file labels.csv --lambda 0.1 --delta 1 --num_classes 0 --output_model_file lsvm_model.bin

then, to use that model to predict classes for the dataset test.csv, storing the output predictions in predictions.csv, the following command may be used:

$ mlpack_linear_svm --input_model_file lsvm_model.bin --test_file test.csv --predictions_file predictions.csv

literally that easy, main character energy fr fr 💜

🎀 OPTIONAL INPUT OPTIONS 🎀

one bullet per option, exactly as they come, no skipping, we respect the source 🫶

  • --delta (-d) [double] — margin of difference between correct class and other classes. default value 1. 💫
  • --epochs (-E) [int] — maximum number of full epochs over dataset for psgd. default value 50.
  • --help (-h) [bool] — default help info. your little bestie when you're lost 🙈
  • --info [string] — print help on a specific option. default value ''.
  • --input_model_file (-m) [unknown] — existing model (parameters).
  • --labels_file (-l) [unknown] — a matrix containing labels (0 or 1) for the points in the training set (y).
  • --lambda (-r) [double] — L2-regularization parameter for training. default value 0.0001.
  • --max_iterations (-n) [int] — maximum iterations for optimizer (0 indicates no limit). default value 10000.
  • --no_intercept (-N) [bool] — do not add the intercept term to the model.
  • --num_classes (-c) [int] — number of classes for classification; if unspecified (or 0), the number of classes found in the labels will be used. default value 0.
  • --optimizer (-O) [string] — optimizer to use for training ('lbfgs' or 'psgd'). default value 'lbfgs'.
  • --seed (-s) [int] — random seed. if 0, std::time(NULL) is used. default value 0.
  • --shuffle (-S) [bool] — don't shuffle the order in which data points are visited for parallel SGD.
  • --step_size (-a) [double] — step size for parallel SGD optimizer. default value 0.01.
  • --test_file (-T) [unknown] — matrix containing test dataset.
  • --test_labels_file (-L) [unknown] — matrix containing test labels.
  • --tolerance (-e) [double] — convergence tolerance for optimizer. default value 1e-10.
  • --training_file (-t) [unknown] — a matrix containing the training set (the matrix of predictors, X).
  • --verbose (-v) [bool] — display informational messages and the full list of parameters and timers at the end of execution.
  • --version (-V) [bool] — display the version of mlpack.

💅 OPTIONAL OUTPUT OPTIONS 💅

the outputs, the moment we've all been waiting for:

  • --output_model_file (-M) [unknown] — output for trained linear svm model. ✨
  • --predictions_file (-P) [unknown] — if test data is specified, this matrix is where the predictions for the test set will be saved.
  • --probabilities_file (-p) [unknown] — if test data is specified, this matrix is where the class probabilities for the test set will be saved.

🧚 ADDITIONAL INFORMATION 🧚

for further information, including relevant papers, citations, and theory, consult the documentation found at http://www.mlpack.org or included with your distribution of mlpack. go read it, it's giving scholar 🦋

and that's the whole thing bestie! this page came from mlpack-4.8.0, dated 22 June 2026, and yeah it's the mlpack_linear_svm(1) manual. periodt 💖 stay regularized, stay convergent, bye!! 🌸✨

📦 where this page comes from