Some novel classifiers designed using prototypes extracted by a new scheme based on self-organizing feature map

Abstract

We propose two new comprehensive schemes for designing prototype-based classifiers. The scheme addresses all major issues (number of prototypes, generation of prototypes, and utilization of the prototypes) involved in the design of a prototype-based classifier. First we use Kohonen's self-organizing feature map (SOFM) algorithm to produce a minimum number (equal to the number of classes) of initial prototypes. Then we use a dynamic prototype generation and tuning algorithm (DYNAGEN) involving merging, splitting, deleting, and retraining of the prototypes to generate an adequate number of useful prototypes. These prototypes are used to design a "1 nearest multiple prototype (1-NMP)" classifier. Though the classifier performs quite well, it cannot reasonably deal with large variation of variance among the data from different classes. To overcome this deficiency we design a "1 most similar prototype (1-MSP)" classifier. We use the prototypes generated by the SOFM-based DYNAGEN algorithm and associate with each of them a zone of influence. A norm (Euclidean)-induced similarity measure is used for this. The prototypes and their zones of influence are fine-tuned by minimizing an error function. Both classifiers are trained and tested using several data sets, and a consistent improvement in performance of the latter over the former has been observed. We also compared our classifiers with some benchmark results available in the literature.

Constructing Compact Takagi-Sugeno Rule Systems: Identification of Complex Interactions in Epidemiological Data

“…Currently, the SVD-QR with column pivoting algorithm has been proposed to perform rule selection for a parsimonious fuzzy rule-base [17]–[19]. Unfortunately, some existing studies have shown great sensitivity to the chosen effective matrix rank (MR) values, so that different estimates of the MR often produce dramatically different rule-reduction results [19].…”

Land cover classification using fuzzy rules and aggregation of contextual information through evidence theory

“…The set of gray values corresponding to a pixel in the channel images is used as the feature vector for that pixel. In the first stage a set of labeled prototypes representing the distribution of the training data is generated using a Self-organizing Feature Map (SOFM) [23] based algorithm developed in [24]. The algorithm dynamically decides the number of prototypes based on the training data.…”

Designing fuzzy rule based classifier using self‐organizing feature map for analysis of multispectral satellite images

“…We use the prototype refinement scheme described in [12,13]. The basic idea behind this refinement algorithm is that a useful prototype, v i should satisfy two criteria:…”

“…After a few iterations this algorithm produces a set of adequate number of prototypes that represents the training data much better than the initial one. For details the readers are referred to [12,13]. Now we use these prototypes to generate fuzzy rules that we describe next.…”