Abstract:Classification of different mechanisms of angle closure glaucoma (ACG) is important for medical diagnosis. Error-correcting output code (ECOC) is an effective approach for multiclass classification. In this study, we propose a new ensemble learning method based on ECOC with application to classification of four ACG mechanisms. The dichotomizers in ECOC are first optimized individually to increase their accuracy and diversity (or interdependence) which is beneficial to the ECOC framework. Specifically, the best feature set is determined for each possible dichotomizer and a wrapper approach is applied to evaluate the classification accuracy of each dichotomizer on the training dataset using cross-validation. The separability of the ECOC codes is maximized by selecting a set of competitive dichotomizers according to a new criterion, in which a regularization term is introduced in consideration of the binary classification performance of each selected dichotomizer. The proposed method is experimentally applied for classifying four ACG mechanisms. The eye images of 152 glaucoma patients are collected by using anterior segment optical coherence tomography (AS-OCT) and then segmented, from which 84 features are extracted. The weighted average classification accuracy of the proposed method is 87.65 % based on the results of leave-one-out cross-validation (LOOCV), which is much better than that of the other existing ECOC methods. The proposed method achieves accurate classification of four ACG mechanisms which is promising to be applied in diagnosis of glaucoma.

Spectral Error Correcting Output Codes for Efficient Multiclass Recognition

An Improved Error-Correcting Output Coding Framework with Kernel-Based Decoding

Adaptive Error-Correcting Output Codes

Distance Function Learning in Error-Correcting Output Coding Framework

Robust spectral regression for face recognition

A novel soft-coded error-correcting output codes algorithm

Heuristic Ternary Error-Correcting Output Codes Via Weight Optimization and Layered Clustering-Based Approach

Constructing the shortest ECOC for fast multi-classification

On Hadamard-Type Output Coding in Multiclass Learning.

The Minimum Closed Sphere Model for Multi-Class Classifiers Using Error-Correcting Output Codes

Learning ECOC Code Matrix for Multiclass Classification with Application to Glaucoma Diagnosis

Using Hybrid Hadamard Error Correcting Output Codes for Multi-class Problem Based on Support Vector Machines

Error-correcting Output Codes Based Ensemble Feature Extraction

Active Learning with Error-Correcting Output Codes

A Dynamic Ensemble Selection Strategy for Improving Error Correcting Output Codes Algorithm.

Joint learning of error-correcting output codes and dichotomizers from data

A New ECOC Algorithm for Multiclass Microarray Data Classification

Robust Multi-View Face Detection Using Error Correcting Output Codes

A Multiclass SVM Method Via Probabilistic Error-Correcting Output Codes

Classification of Multi-Spectral/Hyperspectral Data Using Genetic Programming and Error-Correcting Output Codes

Using Hadamard ECOC in Multi-Class Problems Based on SVM.