Welcome to LightTwinSVM’s documentation!

You can read API documentation, if you want to use the LightTwinSVM’s estimators and tools directly in your project. Furthermore, if you have a question or problem regarding this program, please see the support section of the project in its GitHub repo.

API documentation

This page contains the list of the project’s modules

twinsvm Classes and functios are defined for training and testing TwinSVM classifier.
eval_classifier In this module, classes and methods are defined for evluating the performance of the TwinSVM model.
dataproc In this module, functions for reading and pre-processing datasets are defined.
misc In this module, several miscellaneous functions are defined for using in other module, such as date time formatting and customized progress bar.

Classification results

The LightTwinSVM program saves classification results in an Excel file. Here, the description of each column of the Excel file for both binary and multi-class classification problems is given in below tables. It helps you analyze the classification results.

The description of the Excel file for a binary classification problem.
Column name Description
accuracy The overall classification accuracy
acc_std The standard deviation of the overall classification accuracy
recall_p The recall for the positive class
r_p_std The standard deviation of the recall for the positive class
precision_p The precision for the positive class
p_p_std The standard deviation of the precision for the positive class
f1_p The F1-measure for the positive class
f1_p_std The standard deviation of the F1-measure for the positive class
recall_n The recall for the negative class
r_n_std The standard deviation of the recall for the negative class
precision_n The precision for the negative class
p_n_std The standard deviation of the precision for the negative class
f1_n The F1-measure for the negative class
f1_n_std The standard deviation of the F1-measure for the negative class
tp True positive
tn True negative
fp False positive
fn False negative
C1 The value of the first penalty parameter for TwinSVM
C2 The value of the second penalty parameter for TwinSVM
gamma The value of RBF kernel’s parameter
The description of the Excel file for a multi-class classification problem.
Column name Description
accuracy The overall classification accuracy
acc_std The standard deviation of the overall classification accuracy
micro_recall The micro-averaged recall for all classes
m_rec_std The standard deviation of the micro-averaged recall for all classes
micro_precision The micro-averaged precision for all classes
m_prec_std The standard deviation of the micro-averaged precision for all classes
mirco_f1 The micro-averaged F1-measure for all classes
m_f1_std The standard deviation of the micro-averaged F1-measure for all classes
C1 The value of the first penalty parameter for TwinSVM
C2 The value of the second penalty parameter for TwinSVM
gamma The value of RBF kernel’s parameter

Indices and tables