Abstract:The multi-class classification algorithms are widely used by many areas such as machine learning and computer vision domains. Nowadays, many literatures described multi-class algorithms, however there are few literature that introduced them with thorough theoretical analysis and experimental comparisons. This paper discusses the principles, important parameters, application domain, runtime performance, accuracy, and etc. of twelve multi-class algorithms: decision tree, random forests, extremely randomized trees, multi-class adaboost classifier, stochastic gradient boosting, linear and nonlinear support vector machines, K nearest neighbors, multi-class logistic classifier, multi-layer perceptron, naive Bayesian classifier and conditional random fields classifier. The experiment tested on five data sets: SPECTF heart data set, Ionosphere radar data set, spam junk mail filter data set, optdigits handwriting data set and scene 15 image classification data set. Our major contribution is that we study the relationships between each classifier and impact of each parameters to classification results. The experiment shows that gradient boosted trees, nonlinear support vector machine, K nearest neighbor reach high performance under the circumstance of binary classification and minor data capacity; Under the condition of high dimension, multi-class and big data, however, gradient boosted trees, linear support vector machine, multi-class logistic classifier get good results. At last, the paper addresses the development and future of multi-class classifier algorithms.

Multi-class Image Classification Based on Fast Stochastic Gradient Boosting.

AdaBoost-based Multi-attribute Classification Technology and Its Application

Image Classification Method by Combining Multi-features and Sparse Coding

Multi-class AdaBoost ELM

Stochastic Boosting for Large-scale Image Classification

Multiclass AdaBoost ELM and Its Application in LBP Based Face Recognition

A MKL-MKB Image Classification Algorithm Based on Multi-kernel Boosting Method.

Image Classification Based on KPCA and SVM with Randomized Hyper-parameter Optimization

Discriminative Fast Hierarchical Learning for Multiclass Image Classification

Experimental Comparisons of Multi-class Classifiers.

Automatic image annotation by incorporating feature hierarchy and boosting to scale up SVM classifiers.

Face Recognition Based on Multi-Adaboost

Image based regression using boosting method

Boosting Classifiers for Scene Category Recognition.

Image Classification Using Boosted Local Features with Random Orientation and Location Selection.

An Improved XGBoost Based on Weighted Column Subsampling for Object Classification.

Object Category Recognition Using Boosting Tree With Heterogeneous Features

Multi-class Image Classification Using Deep Learning Algorithm

Ensemble of Fast Learning Stochastic Gradient Boosting.

Neural network for multi-class classification by boosting composite stumps.

Hyperspectral image classification with multiple kernel Boosting algorithm