Previous [ 1] [ 2] [ 3] [ 4] [ 5] [ 6] [ 7] [ 8] [ 9] [ 10] [ 11]

@

Journal of Information Science and Engineering, Vol. 29 No. 4, pp. 765-776 (July 2013)


Improvement of Learning Algorithm for the Multi-instance Multi-label RBF Neural Networks Trained with Imbalanced Samples*


CUNHE LI AND GUOQIANG SHI
College of Computer and Communication Engineering
China University of Petroleum
Qingdao, 266555 P.R. China

Multi-instance multi-label learning (MIML) is a novel learning framework where each sample is represented by multiple instances and associated with multiple class labels. In several learning situations, the multi-instance multi-label RBF neural networks (MIMLRBF) can exploit connections between the instances and the labels of an MIML example directly. However, it is quite often that the numbers of samples in different categories are discrete, i.e., the class distribution is imbalanced. When an MIMLRBF is trained with imbalanced samples, it will produce poor performance for setting the consistent fraction parameter "alpha" for all classes. This paper presents an improved approach in learning algorithms used for training MIMLRBF with imbalanced samples. In the first cluster stage, the methodology calculates the initial medoids for each category based on the data density. Afterwards, k-medoids is been invoked to optimize the medoids. The network will take advantage of the well-adjusted units. In the second stage, the weights between the first and second layer are optimized by the singular value decomposition method. The improved approaches could be used in applications with imbalanced samples. Comparing results employing diverse learning strategies shows interesting outcomes as have come out of this paper.

Keywords: machine learning, radial basis function, multi-instance multi-label learning, class imbalance, neural networks

Full Text () Retrieve PDF document (201307_10.pdf)

Received February 11, 2011; revised December 24, 2011 & February 7, 2012; accepted April 12, 2012.
Communicated by Chih-Jen Lin.
* This work is supported by the Fundamental Research Funds for the Central Universities (No. 09CX04031A).