Improving naive Bayes classifiers using neuro-fuzzy learning

Nürnberger, Andreas; Borgelt, Christian; Klose, Aljoscha

doi:10.1109/iconip.1999.843978

Cited by 17 publications

(9 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When we use the single winner method in (15), the estimated class boundary is calculated as by the first definition and by the second definition from the equation (28) The estimated class boundary has a large error in the case of the first definition while it is almost the same as the actual threshold in the second definition. The large error by the first definition is due to the large rule weight of the fuzzy rule .…”

Section: A Simulation Results On Single-dimensional Problemsmentioning

confidence: 99%

“…If multiple fuzzy rules have the same maximum value but different consequent classes for the new pattern in (15), the classification of is rejected. The classification is also rejected if no fuzzy rule is compatible with the new pattern .…”

Section: Fuzzy Reasoning For Pattern Classificationmentioning

confidence: 99%

“…Thus the rule weight can be viewed as adjusting the compatibility grade (i.e., applicability) of each fuzzy rule to the current input vector. When there is exactly one rule per class in neuro-fuzzy naive Bayes classifiers [15], the rule weight and the compatibility grade are explained as being the prior probability of the consequent class and the probability density function, respectively. Under this interpretation, the multiplication of the compatibility grade by the rule weight becomes natural.…”

Section: Fuzzy Reasoning For Pattern Classificationmentioning

confidence: 99%

See 2 more Smart Citations

Rule weight specification in fuzzy rule-based classification systems

Ishibuchi¹,

Yamamoto²

2005

IEEE Trans. Fuzzy Syst.

425

213

View full text Add to dashboard Cite

Abstract-This paper shows how the rule weight of each fuzzy rule can be specified in fuzzy rule-based classification systems. First, we propose two heuristic methods for rule weight specification. Next, the proposed methods are compared with existing ones through computer simulations on artificial numerical examples and real-world pattern classification problems. Simulation results show that the proposed methods outperform the existing ones in the case of multiclass pattern classification problems with many classes.

show abstract

Section: A Simulation Results On Single-dimensional Problemsmentioning

confidence: 99%

Section: Fuzzy Reasoning For Pattern Classificationmentioning

confidence: 99%

Section: Fuzzy Reasoning For Pattern Classificationmentioning

confidence: 99%

See 1 more Smart Citation

Rule weight specification in fuzzy rule-based classification systems

Ishibuchi¹,

Yamamoto²

2005

IEEE Trans. Fuzzy Syst.

425

213

View full text Add to dashboard Cite

show abstract

“…chapter 8). Furthermore there are some interesting connections to fuzzy clustering [Borgelt et al 1999] and neuro-fuzzy rule induction [Nürnberger et al 1999] through naive Bayes classifiers, which may lead to powerful hybrid systems.…”

Section: Graphical Modelsmentioning

confidence: 99%

Data Mining with Graphical Models

Kruse

Borgelt

2002

Lecture Notes in Computer Science

View full text Add to dashboard Cite

“…Furthermore, we give a formula description of the shape of the class borders and we extend our discussion to the influence of rule weights, which are used in a number of fuzzy classification systems. Based on these considerations, we also point out some aspects of the classification behavior of naive Bayes classifiers, since they can be seen as a subset of fuzzy systems, as we have shown in [8].…”

Section: Introductionmentioning

confidence: 95%