Comparison and design of interpretable linguistic vs. scatter FRBSs: Gm3m generalization and new rule meaning index for global assessment and local pseudo-linguistic representation

Galende, Marta; Gacto, María José; Sainz, Gregorio; Alcalá, Rafael

doi:10.1016/j.ins.2014.05.023

Cited by 12 publications

(13 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Secondly, we have those involving semantic issues such as: the distinguishability of the fuzzy sets (Oliveira, 1999), consistence and similarity of the fuzzy rule base (Alonso and Magdalena, 2011;Pulkkinen et al, 2008), the number of rules simultaneously fired (Pancho et al, 2013;Márquez et al, 2012), etc. ; or those defining more complex indexes to obtain some semantic restrictions, such as GM3M (Gacto et al, 2010), RBC (Alonso and Magdalena, 2011), Integrity I (Antonelli et al, 2011), Transparency (Pulkkinen et al, 2008), Cointension (Mencar et al, 2011) or RMI (Galende et al, 2014).…”

Section: Accuracy-interpretability Trade-offmentioning

confidence: 99%

“…Another view is based on an existing, improved FRBS: approaches based on rule selection and MOEA are developed in Galende et al (2012); Márquez et al (2012); Pulkkinen et al (2008); Ishibuchi and Nojima (2007); Ishibuchi et al (2001) and Ishibuchi et al (1997). A GA tuning based approach can be found in Roubos and Setnes (2001), while MOEA based rule selection and tuning is used by Galende et al (2014); Fazzolari et al (2013b); Alcalá et al (2011) and Gacto et al (2010). This view is used in this approach.…”

Section: Accuracy-interpretability Trade-offmentioning

confidence: 99%

See 1 more Smart Citation

Multi-objective based Fuzzy Rule Based Systems (FRBSs) for trade-off improvement in accuracy and interpretability: A rule relevance point of view.

Rey¹,

Galende²,

Fuente

et al. 2017

Knowledge-Based Systems

View full text Add to dashboard Cite

Fuzzy rule-based systems (FRBSs) are a common alternative for applying fuzzy logic in different areas and real-world problems. The schemes and algorithms used to generate these types of systems imply that their performance can be analyzed from different points of view, not only model accuracy. Any model, including fuzzy models, needs to be sufficiently accurate, but other perspectives, such as interpretability, are also possible for the FRBSs. Thus, the Accuracy-Interpretability trade-off arises as a challenge for fuzzy systems, as approaches are currently able to generate FRBSs with different trade-offs. Here, rule Relevance is added to Accuracy and Interpretability for a better trade-off in FRBSs. These three factors are involved in this approach to make a rule selection using a multi-objective evolutionary algorithm. The proposal has been tested and compared with nine datasets, two linguistic and two scatter fuzzy algorithms, four measures of interpretability and two rule relevance formulations. The results have been analyzed for different views of Interpretability, Accuracy and Relevance, and the statistical tests have shown that significant improvements have been achieved. On the other hand, the Relevance-based role of fuzzy rules has been checked, and it has been shown that low Relevance rules have a relevant role for trade-off, $ This work has been partially supported by the Spanish Ministry of Economy and Competitiveness through the Project no. DPI2015-67341-C2-2-R

show abstract

Section: Accuracy-interpretability Trade-offmentioning

confidence: 99%

Section: Accuracy-interpretability Trade-offmentioning

confidence: 99%

Multi-objective based Fuzzy Rule Based Systems (FRBSs) for trade-off improvement in accuracy and interpretability: A rule relevance point of view.

Rey¹,

Galende²,

Fuente

et al. 2017

Knowledge-Based Systems

View full text Add to dashboard Cite

show abstract

“…The classification rate and the error rate of S are often used as the accuracy measures, while the number of rules and the number of conditions in S are often used as the complexity measures. Well-designed interpretability measures are also used instead of the complexity measures in some studies [14], [15]. In recent years, multiobjective GFS (MoGFS) have attracted increasing attention in the research field of GFS [1]- [3], [7], [12], [23].…”

Section: Fitness(s) = W 1 Accuracy(s) − W 2 Complexity(s)mentioning

confidence: 99%

Application of Parallel Distributed Implementation to Multiobjective Fuzzy Genetics-Based Machine Learning

Nojima

Takahashi

Ishibuchi

2015

Intelligent Information and Database Systems

View full text Add to dashboard Cite

Abstract. Fuzzy genetics-based machine learning is one of data mining techniques based on evolutionary computation. It can generate accurate classifiers with a small number of fuzzy if-then rules from numerical data. Its multiobjective version can provide a number of classifiers with a different tradeoff between accuracy and complexity. One major drawback of this method is the computation time when we use it for large data sets. In our previous study, we proposed parallel distributed implementation of single-objective fuzzy genetics-based machine learning which could drastically reduce the computation time. In this paper, we apply our idea of parallel distributed implementation to multiobjective fuzzy genetics-based machine learning. Through computational experiments on large data sets, we examine the effects of parallel distributed implementation on the search performance of our multiobjective fuzzy genetics-based machine learning and its computation time.

show abstract

“…Our aim is to propose a new extension of the wellknown linguistic Fuzzy Rule-Based Systems (FRBS) [19][20][21] that is specifically designed for regression problems. This is done by combining new linguistic fuzzy grammar and a novel interpretable linear extension in order to obtain the highest possible transparency level, as this is our main objective, as well as a rule set of no more than 6 or 7 rules [22] for regression problems.…”

Section: Introductionmentioning

confidence: 99%

“…Moreover, we propose an enhanced multiobjective evolutionary algorithm (MOEA) in two stages (learning linguistic partitions and rules, plus tuning and rule selection) to optimize accuracy together with some well-known interpretability measures from the specialized literature that account for the number of rules and the overlap in the linguistic terms and/or rule inconsistency (the GM3M and RMI indexes [19,21]). The main contribution in this algorithmic part is the inclusion of a new linguistic tree-based Rule Base (RB) learning algorithm that adapts perfectly to the new type of rules and therefore enhances the said evolutionary learning method, as node conditions can be fuzzified and the linear extension can be obtained at the tree leaves.…”

Section: Introductionmentioning

confidence: 99%

Transparent but Accurate Evolutionary Regression Combining New Linguistic Fuzzy Grammar and a Novel Interpretable Linear Extension

Biedma-Rdguez

Gacto

Anguita‐Ruiz

et al. 2022

Int. J. Fuzzy Syst.

Self Cite

View full text Add to dashboard Cite

Scientists must understand what machines do (systems should not behave like a black box), because in many cases how they predict is more important than what they predict. In this work, we propose a new extension of the fuzzy linguistic grammar and a mainly novel interpretable linear extension for regression problems, together with an enhanced new linguistic tree-based evolutionary multiobjective learning approach. This allows the general behavior of the data covered, as well as their specific variability, to be expressed as a single rule. In order to ensure the highest transparency and accuracy values, this learning process maximizes two widely accepted semantic metrics and also minimizes both the number of rules and the model mean squared error. The results obtained in 23 regression datasets show the effectiveness of the proposed method by applying statistical tests to the said metrics, which cover the different aspects of the interpretability of linguistic fuzzy models. This learning process has obtained the preservation of high-level semantics and less than 5 rules on average, while it still clearly outperforms some of the previous state-of-the-art linguistic fuzzy regression methods for learning interpretable regression linguistic fuzzy systems, and even to a competitive, pure accuracy-oriented linguistic learning approach. Finally, we analyze a case study in a real problem related to childhood obesity, and a real expert carries out the analysis shown.

show abstract

Comparison and design of interpretable linguistic vs. scatter FRBSs: Gm3m generalization and new rule meaning index for global assessment and local pseudo-linguistic representation

Cited by 12 publications

References 47 publications

Multi-objective based Fuzzy Rule Based Systems (FRBSs) for trade-off improvement in accuracy and interpretability: A rule relevance point of view.

Multi-objective based Fuzzy Rule Based Systems (FRBSs) for trade-off improvement in accuracy and interpretability: A rule relevance point of view.

Application of Parallel Distributed Implementation to Multiobjective Fuzzy Genetics-Based Machine Learning

Transparent but Accurate Evolutionary Regression Combining New Linguistic Fuzzy Grammar and a Novel Interpretable Linear Extension

Contact Info

Product

Resources

About