Boosting methods for multi-class imbalanced data classification: an experimental review (2020-09-01T00:00:00.000000Z)

TL;DR

The experimental studies show that the CatBoost and LogitBoost algorithms are superior to other boosting algorithms on multi- class imbalanced conventional and big datasets, respectively and the MMCC is a better evaluation metric than the MAUC and G-mean in multi-class imbalanced data domains.

Abstract

Since canonical machine learning algorithms assume that the dataset has equal number of samples in each class, binary classification became a very challenging task to discriminate the minority class samples efficiently in imbalanced datasets. For this reason, researchers have been paid attention and have proposed many methods to deal with this problem, which can be broadly categorized into data level and algorithm level. Besides, multi-class imbalanced learning is much harder than binary one and is still an open problem. Boosting algorithms are a class of ensemble learning methods in machine learning that improves the performance of separate base learners by combining them into a composite whole. This paper’s aim is to review the most significant published boosting techniques on multi-class imbalanced datasets. A thorough empirical comparison is conducted to analyze the performance of binary and multi-class boosting algorithms on various multi-class imbalanced datasets. In addition, based on the obtained results for performance evaluation metrics and a recently proposed criteria for comparing metrics, the selected metrics are compared to determine a suitable performance metric for multi-class imbalanced datasets. The experimental studies show that the CatBoost and LogitBoost algorithms are superior to other boosting algorithms on multi-class imbalanced conventional and big datasets, respectively. Furthermore, the MMCC is a better evaluation metric than the MAUC and G-mean in multi-class imbalanced data domains.

Authors

J. Tanha

1 papers

Yousef Abdi

1 papers

Negin Samadi

1 papers

TL;DR

Abstract

Authors

References86 items

Using SMOTE to Deal with Class-Imbalance Problem in Bioactivity Data to Predict mTOR Inhibitors

Techniques to deal with imbalanced data in multi-class problems: A review of existing methods

Time-aware domain-based social influence prediction

The advantages of the Matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation

Bagging Using Instance-Level Difficulty for Multi-Class Imbalanced Big Data Classification on Spark

Sampling Techniques to Overcome Class Imbalance in a Cyberbullying Context

A Classification Model For Class Imbalance Dataset Using Genetic Programming

Empirical Comparison of Area under ROC curve (AUC) and Mathew Correlation Coefficient (MCC) for Evaluating Machine Learning Algorithms on Imbalanced Datasets for Binary Classification

TLUSBoost algorithm: a boosting solution for class imbalance problem

A survey on addressing high-class imbalance in big data

An empirical comparison on state-of-the-art multi-class imbalance learning algorithms and a new diversified ensemble learning scheme

LIUBoost: Locality Informed Under-Boosting for Imbalanced Data Classification

Class imbalance ensemble learning based on the margin theory

LightGBM: A Highly Efficient Gradient Boosting Decision Tree

MEBoost: Mixing estimators with boosting for imbalanced data classification

isGPT: An optimized model to identify sub-Golgi protein types using SVM and Random Forest based feature selection

CatBoost: unbiased boosting with categorical features

Exploring the effectiveness of dynamic ensemble selection in the one-versus-one scheme

BVDT: A Boosted Vector Decision Tree Algorithm for Multi-Class Classification Problems

Empowering one-vs-one decomposition with ensemble learning for multi-class imbalanced data

An improved method to construct basic probability assignment based on the confusion matrix for classification problem

Learning from imbalanced data: open challenges and future directions

XGBoost: A Scalable Tree Boosting System

A hybrid one-class rule learning approach based on swarm intelligence for software fault prediction

A Survey on Methods for Solving Data Imbalance Problem for Classification

Random Balance: Ensembles of variable priors classifiers for imbalanced data

The Precision-Recall Plot Is More Informative than the ROC Plot When Evaluating Binary Classifiers on Imbalanced Datasets

Neighbourhood sampling in bagging for imbalanced data

Prediction of human breast and colon cancers from imbalanced data using nearest neighbor and support vector machines

An improved multiclass LogitBoost using adaptive-one-vs-one

EUSBoost: Enhancing ensembles for highly imbalanced data-sets by evolutionary undersampling

Analysing the classification of imbalanced data-sets with multiple classes: Binarization techniques and ad-hoc approaches

SMOTE-RSB*: a hybrid preprocessing approach based on oversampling and undersampling for high imbalanced data-sets using SMOTE and rough sets theory

Multiclass Imbalance Problems: Analysis and Potential Solutions

A Review on Ensembles for the Class Imbalance Problem: Bagging-, Boosting-, and Hybrid-Based Approaches

Improving software fault-prediction for imbalanced data

Multiclass Boosting: Theory and Algorithms

AOSO-LogitBoost: Adaptive One-Vs-One LogitBoost for Multi-Class Problem

RAMOBoost: Ranked Minority Oversampling in Boosting

Multi-class AdaBoost with Hypothesis Margin

A Comparison of MCC and CEN Error Measures in Multi-Class Prediction

Ensemble-based classifiers

ABC-LogitBoost for Multi-class Classification

Handling Class Imbalance Problem in Cultural Modeling

Cost-sensitive boosting for classification of imbalanced data

Boosted Classification Trees and Class Probability/Quantile Estimation

Boosting for Learning Multiple Classes with Imbalanced Class Distribution

Statistical Comparisons of Classifiers over Multiple Data Sets

Using AUC and accuracy in evaluating learning algorithms

A study of the behavior of several methods for balancing machine learning training data

SMOTEBoost: Improving Prediction of the Minority Class in Boosting

AUC: a Statistically Consistent and more Discriminating Measure than Accuracy

Predicting rare classes: can boosting make any weak learner strong?

Data mining and knowledge discovery in proton nuclear magnetic resonance (1H-NMR) spectra using frequency to information transformation (FIT)

Evaluating boosting algorithms to classify rare classes: comparison and improvements

Greedy function approximation: A gradient boosting machine.

A Comparative Study of Cost-Sensitive Boosting Algorithms

MadaBoost: A Modification of AdaBoost

An Empirical Study of MetaCost Using Boosting Algorithms

BoosTexter: A Boosting-based System for Text Categorization

Special Invited Paper-Additive logistic regression: A statistical view of boosting

A Brief Introduction to Boosting

AdaCost: Misclassification Cost-Sensitive Boosting

Machine Learning for the Detection of Oil Spills in Satellite Radar Images

A decision-theoretic generalization of on-line learning and an application to boosting

Bagging Predictors

Proper Bayes Minimax Estimators of the Multivariate Normal Mean Vector for the Case of Common Unknown Variances

can boosting make any weak learner strong? In: Proceedings of the eighth ACM SIGKDD international conference on Knowledge discovery and data mining; 2002

Significance of non-parametric statistical tests for comparison of classifiers over multiple datasets

Combining One-vs-One Decomposition and Ensemble Learning for Multi-class Imbalanced Data

An overview of classification algorithms for imbalanced datasets

A New Feature Selection Method for Internet Traffic Classification Using ML

Boosting Algorithms: A Review of Methods, Theory, and Applications

Improving Electric Fraud Detection using Class Imbalance Strategies

A New approach for Classification of Highly Imbalanced Datasets using Evolutionary Algorithms

RUSBoost: A Hybrid Approach to Alleviating Class Imbalance