The present article provides a survey of the available literature on data mining using soft computing. A categorization has been provided based on the different soft computing tools and their hybridizations used, the data mining function implemented, and the preference criterion selected by the model. The utility of the different soft computing methodologies is highlighted. Generally fuzzy sets are suitable for handling the issues related to understandability of patterns, incomplete/noisy data, mixed media information and human interaction, and can provide approximate solutions faster. Neural networks are nonparametric, robust, and exhibit good learning and generalization capabilities in data-rich environments. Genetic algorithms provide efficient search algorithms to select a model, from mixed media data, based on some preference criterion/objective function. Rough sets are suitable for handling different types of uncertainty in data. Some challenges to data mining and the application of soft computing methodologies are indicated. An extensive bibliography is also included.

There has been an increased interest in combining fuzzy systems with neural networks because fuzzy neural systems merge the advantages of both paradigms. On the one hand, parameters in fuzzy systems have clear physical meanings and rule-based and linguistic information can be incorporated into adaptive fuzzy systems in a systematic way. On the other hand, there exist powerful algorithms for training various neural network models. However, most of the proposed combined architectures are only able to process static input-output relationships, i.e. they are not able to process temporal input sequences of arbitrary length. Fuzzy finite-state automata (FFAs) can model dynamical processes whose current state depends on the current input and previous states. Unlike in the case of deterministic finite-state automata (DFAs), FFAs are not in one particular state, rather each state is occupied to some degree defined by a membership function. Based on previous work on encoding DFAs in discrete-tim...

Learning a large number of simple local concepts is both faster and easier than learning a single global concept. Inspired by this principle of divide and conquer, a number of modular learning approaches have been proposed by the computational intelligence community. In modular learning, the classification/regression/clustering problem is first decomposed into a number of simpler subproblems, a module is learned for each of these subproblems, and finally their results are integrated by a suitable combining method. Mixtures of experts and clustering are two of the techniques that are describable in this paradigm. In this paper we present a broad framework for Generalized Associative Modular Learning Systems (GAMLS). Modularity is introduced through soft association of each training pattern with every module. The coupled problems of learning the module parameters and learning associations are solved iteratively using deterministic annealing. Starting at a high temperature with only one modu...

The knowledge base in expert systems usually contains different types of information which can be classified as explicit and implicit with respect to its representation. The explicit representation is based on a symbolic expression of human expert knowledge while the numerical data which require additional processing to be really understood represent the implicit knowledge. The rule-based systems and neural networks are typical examples of these different representation approaches. The main problem of rule-based systems is the knowledge acquisition which can be overcoming by learning and adaptation in neural networks. On the other hand, the neural implicit knowledge representation loses the capability to explain and justify the inference. Thus, the advantages and disadvantages of explicit and implicit knowledge representation in expert systems are complementary and we will first give a general comparison of both. Then we will discuss how to process the neural knowledge to embed it into...

. This chapter contains a discussion material and preliminary experimental results of a new approach to building intelligent conscious machines (ICM) and its application to multilingual spoken recognition systems. ICM can analyse their behaviour and subsequently adapt and improve their structure and functionality during operation, can evaluate their ability of problem solving in terms of what they "can do" and what they cannot. These systems consist of many modules interacting during operation and organised in several hierarchical levels aggregated into two main ones, a low, sub-conscious level, and a higher, conscious level. A framework for intelligent conscious machines is proposed and a partial realisation is presented which makes use of fuzzy neural networks and spatial-temporal maps. The framework is applicable to recognising patterns from time-series at different time scales, with numerous applications. A particular case study of spoken language recognition is presented along wit...

the paper presented fuzzy logics

A new methodology for developing Expert System (ES) is presented in this paper. It has learning ability to extract knowledge by a simple knowledge base using the learning by examples paradigm. The choice of a simple knowledge base was motivated by the fact that, in this case, it is easier to have consistence in putting together the several pieces of knowledge. So, the problems attached with knowledge elicitation phase are simplified. The

This paper presents results from a research project on speech data analysis and speaker independent adaptive speech recognition using novel neurofuzzy techniques and a new system architecture. A speech recognition system of English is presented in the chapter that utilises fuzzy neural networks and selforganising spatial-temporal maps. Experimental results on speech recognition are given.

The emerging need for Hybrid Intelligent Systems (HIS) is currently motivating important research and development work. The integration of different learning and adaptation techniques, to overcome individual limitations and achieve synergetic effects through hybridization or fusion of these techniques, has in recent years contributed to a large number of new intelligent system designs. Soft Computing (SC) introduced by Lotfi Zadeh [1] is an innovative approach to construct computationally intelligent hybrid systems consisting of Artificial Neural Network (ANN), Fuzzy Logic (FL), approximate reasoning and derivative free optimization methods such as Genetic Algorithm (GA), Simulated Annealing (SA) and Tabu Search (TS). Most of these approaches, however, follow an ad hoc design methodology, further justified by success in certain application domains. Due to the lack of a common framework it remains often difficult to compare the various hybrid systems conceptually and evaluate their performance comparatively. It has been over a decade since HIS were first applied to solve complicated problems. In this paper, we first aim at classifying state--of--the--art intelligent systems, which have evolved over the past decade in the HIS community. Some theoretical concepts of ANN, FL and Global Optimization Algorithms (GOA) namely GA, SA and TS are also presented. We further attempt to summarize the work that has been done and present the current standing of our vision on HIS and future research directions.

This paper introduces a new learning algorithm for artificial neural networks, based on a fuzzy inference system ANBLIR. It is a computationally effective neuro-fuzzy system with parametrized fuzzy sets in the consequent parts of fuzzy if-then rules, which uses a conjunctive as well as a logical interpretation of those rules. In the original approach, the estimation of unknown system parameters was made by means of a combination of both gradient and least-squares methods. The novelty of the learning algorithm consists in the application of a deterministic annealing optimization method. It leads to an improvement in the neuro-fuzzy modelling performance. To show the validity of the introduced method, two examples of application concerning chaotic time series prediction and system identification problems are provided.