PROGRESSIVE LEARNING - MULTI-CLASS CLASSIFIER

Research Aim

To develop human-learning-inspired progressive learning technique (PLT) and incorporate it for multi-class classification problems. The progressive learning technique is the next logical step in the sequence of batch and online learning. This technique is suitable for real-world applications where the number of classes is often unknown and online learning from real-time data is required. The consistency, complexity, performance and speed of the PLT are evaluated and analyzed. 

Datasets

Iris, Balance-scale, Waveform, Wine, Satellite image, Digit, Character dataset 1, Character dataset 2, Character dataset 3

Dataset specifications are given below:

Algorithm

Code

The source code for progressive learning technique for multi-class classification is available in github (rajasekar-venkatesan/progressive-multiclass)

Results

More detailed discussions and results are available in the paper:

A Novel Progressive Learning Technique for Multi-class Classification [PDF][Code]