Abstract:Monotonic classification problems mean that both feature values and class labels are ordered and monotonicity relationships exist between some features and the decision label. Extreme Learning Machine (ELM) is a single hidden layer feedforward neural network with fast training rate and good generalization capability, but due to the existence of training error, ELM cannot be directly used to handle monotonic classification problems. This work proposes a generalization of ELM for processing the monotonic classification, named as Monotonic Classification Extreme Learning Machine (MCELM) in which the monotonicity constraints are imposed to the original ELM model. Mathematically, MCELM is a quadratic programming problem in which the monotonicity relationships are considered as constraints and the training error is the objective to be minimized. The mathematical model of MCELM not only can make the generated classifier monotonic but also can minimize the classification error. MCELM does not need to tune parameters iteratively, and therefore, keeps the advantage of extremely fast training which is the essential characteristic of ELM. MCELM does not require that the monotonic relationships existing between features and the output are consistent, which essentially relaxes the assumption of consistent monotonicity used in most existing approaches to handling monotonic classification problems. In comparison with exiting approaches to handling monotonic classification, MCELM can indeed generate a monotonicity-reserving classifier which experimentally shows a much better generalization capability on both artificial and real world datasets.

Text Categorization Based on Regularization Extreme Learning Machine

Collaborative Work with Linear Classifier and Extreme Learning Machine for Fast Text Categorization

A Grouped Structure-based Regularized Regression Model for Text Categorization

Correntropy Extreme Learning Machine Based on Spatial Pyramid Matching and Local Receptive Field

Aggressive Dimensionality Reduction With Reinforcement Local Feature Selection For Text Categorization

Rough Extreme Learning Machine: a New Classification Method Based on Uncertainty Measure

Embedding Regularizer Learning for Multi-View Semi-Supervised Classification.

Image Classification Using Low-Rank Regularized Extreme Learning Machine

Research on Extreme Learning of Neural Networks

Multilayer discriminative extreme learning machine for classification

Comparing effectiveness of regularization methods on text classification: Simple and complex model in data shortage situation

Lagrangian supervised and semi-supervised extreme learning machine

Robust Multi-Layer Extreme Learning Machine Using Bias-Variance Tradeoff

A Smooth Extreme Learning Machine for Classification

Continual Learning for Text Classification with Information Disentanglement Based Regularization

Regularized Correntropy Criterion Based Semi-Supervised ELM.

Regularized Minimum Class Variance Extreme Learning Machine for Language Recognition

Multi-Label Learning with Regularization Enriched Label-Specific Features

A new regularization classification method based on extreme learning machine in network data

Monotonic classification extreme learning machine.

Gated Neural Network with Regularized Loss for Multi-label Text Classification.