Rule insertion and rule extraction from evolving fuzzy neural networks: algorithms and applications for building adaptive, intelligent expert systems

Kasabov, Nikola; Woodford, Brendon J.

doi:10.1109/fuzzy.1999.790109

Cited by 45 publications

(13 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Seven major requirements of the present ISs (that are addressed in the ECOS framework presented in [39] and [43]) are discussed in [37], [39], [40], and [43]. They are concerned with fast learning, online incremental adaptive learning, open structure organization, memorising information, active interaction, Manuscript received June 13, 2001; revised August 6, 2001.…”

Section: Introductionmentioning

confidence: 99%

DENFIS: dynamic evolving neural-fuzzy inference system and its application for time-series prediction

Kasabov

Song

2002

IEEE Trans. Fuzzy Syst.

1,140

597

View full text Add to dashboard Cite

Abstract-This paper introduces a new type of fuzzy inference systems, denoted as dynamic evolving neural-fuzzy inference system (DENFIS), for adaptive online and offline learning, and their application for dynamic time series prediction. DENFIS evolve through incremental, hybrid (supervised/unsupervised), learning, and accommodate new input data, including new features, new classes, etc., through local element tuning. New fuzzy rules are created and updated during the operation of the system. At each time moment, the output of DENFIS is calculated through a fuzzy inference system based on -most activated fuzzy rules which are dynamically chosen from a fuzzy rule set. Two approaches are proposed: 1) dynamic creation of a first-order Takagi-Sugeno-type fuzzy rule set for a DENFIS online model; and 2) creation of a first-order Takagi-Sugeno-type fuzzy rule set, or an expanded high-order one, for a DENFIS offline model. A set of fuzzy rules can be inserted into DENFIS before or during its learning process. Fuzzy rules can also be extracted during or after the learning process. An evolving clustering method (ECM), which is employed in both online and offline DENFIS models, is also introduced. It is demonstrated that DENFIS can effectively learn complex temporal sequences in an adaptive way and outperform some well-known, existing models.Index Terms-Dynamic evolving neural-fuzzy inference system (DENFIS), hybrid systems, online adaptive learning, online clustering, time series prediction.

show abstract

Section: Introductionmentioning

confidence: 99%

DENFIS: dynamic evolving neural-fuzzy inference system and its application for time-series prediction

Kasabov

Song

2002

IEEE Trans. Fuzzy Syst.

1,140

597

View full text Add to dashboard Cite

show abstract

“…As far as the classification rate and the number of rules are concerned, our method converges to a network of just three rules, no more than the classes of the problem, while not loosing in performance when compared to networks with larger numbers of hidden nodes. In addition to the experiments presented herein, our method outperforms others in the literature as well (7 rules, 96.7% [5]) (17 rules, 95.3% [6]) (9 rules, 95.3% [7]) (7 rules, 96% [8]). …”

Section: Resultsmentioning

confidence: 79%

Improving the Performance of Resource Allocation Networks through Hierarchical Clustering of High-Dimensional Data

Tsapatsoulis

Wallace

Kasderidis

2003

Artificial Neural Networks and Neural Information Processing — ICANN/ICONIP 2003

View full text Add to dashboard Cite

Abstract. Adaptivity to non-stationary contexts is a very important property for intelligent systems in general, as well as to a variety of applications of knowledge based systems in era of "ambient intelligence". In this paper we present a modified Resource Allocating Network architecture that allows for online adaptation and knowledge modelling through its adaptive structure. As in any neural network system proper parameter initialization reduces training time and effort. However, in RAN architectures, proper parameter initialization also leads to compact modelling (less hidden nodes) of the process under examination, and consequently to better generalization. In the cases of high-dimensional data parameter initialization is both difficult and time consuming. In the proposed scheme a high -dimensional, unsupervised clustering method is used to properly initialize the RAN architecture. Clusters correspond to the initial nodes of RAN, while output layer weights are also extracted from the clustering procedure. The efficiency of the proposed method has been tested on several classes of publicly available data (iris, ionosphere, etc.)

show abstract

“…1 which provides order steps of how to distinguish between phishing emails and ham emails. This framework divided to four stages, first stage pre-processing of the data set, second stage is email object similarity and third stage is integrated with Evolving Fuzzy Neural Networks (Kasabov and Woodford, 1999) to build PENF, for detection and prediction phishing emails in online mode. All of this stages will work after determine the features of phishing email which used in our framework.…”

Section: Methodsmentioning

confidence: 99%

“…The expert systems EFuNN, is an intelligent system that evolves in accordance with the Evolving Connectionist System (ECOS) (Kasabov and Woodford, 1999). This system, like traditional expert systems with more power full included working with unfixed number of rules used to develop the Artificial Intelligent (AI).…”

Section: Adoptive Evolving Fuzzy Neural Network (Efunn)mentioning

confidence: 99%

“…In this study we try to detect and predict zero-day" phishing Emails by a new framework called phishing Evolving Neural Fuzzy Framework (PENFF) based on adoptive Evolving Fuzzy Neural Networks (EFuNN) (Kasabov and Woodford, 1999), PENFF can be distinguish between phishing email and ham(legitimate) email in online mode footprint consuming memory ,increase the level of accuracy in prediction process and has ability to improve the rule continually, This study devoted concentration on phishing email only. The remainder of this study is ordered as follows.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Evolving Fuzzy Neural Network for Phishing Emails Detection

Almomani

Wan

Altaher

et al. 2012

Journal of Computer Science

View full text Add to dashboard Cite

Abstract:One of the broadly used internet attacks to deceive customers financially in banks and agencies is unknown "zero-day" phishing Emails "zero-day" phishing Emails is a new phishing email that it has not been trained on old dataset, not included in black list. Accordingly, the current paper seeks to Detection and Prediction of unknown "zero-day" phishing Emails by provide a new framework called Phishing Evolving Neural Fuzzy Framework (PENFF) that is based on adoptive Evolving Fuzzy Neural Network (EFuNN). PENFF does the process of detection of phishing email depending on the level of features similarity between body email and URL email features. The totality of the common features vector is controlled by EFuNN to create rules that help predict the phishing email value in online mode. The proposed framework has proved its ability to detect phishing emails by decreasing the error rate in classification process. The current approach is considered a highly compacted framework. As a performance indicator; the Root Mean Square Error (RMSE) and Non-Dimensional Error Index (NDEI) has 0.12 and 0.21 respectively, which has low error rate compared with other approaches Furthermore, this approach has learning capability with footprint consuming memory.

show abstract

Rule insertion and rule extraction from evolving fuzzy neural networks: algorithms and applications for building adaptive, intelligent expert systems

Cited by 45 publications

References 8 publications

DENFIS: dynamic evolving neural-fuzzy inference system and its application for time-series prediction

DENFIS: dynamic evolving neural-fuzzy inference system and its application for time-series prediction

Improving the Performance of Resource Allocation Networks through Hierarchical Clustering of High-Dimensional Data

Evolving Fuzzy Neural Network for Phishing Emails Detection

Contact Info

Product

Resources

About