Adnan KhanAdnanProduct & AI
San Francisco Bay Area
Get in touch

Source note2 min read

Interpretable Fuzzy Rule-Based Regression Extension for Ex-Fuzzy Library

Source

arXiv

arxiv.org/abs/2607.20277v1

Read the source

Source

Plain-Language Brief

Researchers present an extension for the Ex-Fuzzy library that makes Mamdani-style fuzzy regression more accessible in modern machine learning software. This extension allows for learning from data and offers transparency, using a target-aware partition initialization strategy based on Fuzzy C-Means clustering. The study compares Gaussian and trapezoidal partitions against standard baseline models, showing that Gaussian partitions outperform uniform trapezoidal partitions and achieve a mean coefficient of determination of approximately 0.86.

What the Document Says

  • The research introduces an extension for the Ex-Fuzzy library, which enables Mamdani fuzzy inference with scalar consequents learned directly from data.
  • A target-aware partition initialization strategy is introduced, based on Fuzzy C-Means clustering.
  • Linguistic variables are derived from an augmented input-output space to emphasise output-relevant regions of the feature space.
  • The extension is evaluated on ten regression datasets.
  • Gaussian partitions outperform uniform trapezoidal partitions in the study.

Key Concepts Mentioned

  • Machine learning
  • Fuzzy rule-based systems
  • Mamdani-style fuzzy regression
  • Ex-Fuzzy library
  • Fuzzy C-Means clustering
  • Regression tasks
  • Safety-critical and regulated domains

Claims Made by the Document

  • The proposed extension offers transparency and competitive predictive performance.

Evidence Used

  • Experimental results show that Gaussian partitions outperform uniform trapezoidal partitions, achieving a mean coefficient of determination of approximately 0.86.

Methods / Approach

  • Mamdani fuzzy inference
  • Fuzzy C-Means clustering
  • Comparison against linear regression, multilayer perceptron, and random forests

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

The study does not discuss the limitations of the proposed extension for real-world applications or its performance on large datasets.

Open Questions

  • How well does the proposed extension perform on larger datasets?
  • What are the real-world implications of using this extension in safety-critical and regulated domains?

Notes for Future Q&A

May include interpretation. Labeled where it does.

The study focuses on evaluating the extension's performance compared to other models, but it does not provide detailed information about its practical implementation or potential limitations in real-world scenarios.

Raw Abstract / Excerpt

Machine learning models achieve high predictive accuracy in regression tasks, but their deployment in safety-critical and regulated domains requires interpretability. While fuzzy rule-based systems offer transparent, linguistically explicit interpretable models, Mamdani-style fuzzy regression remains underrepresented in modern machine learning software libraries. This paper presents an interpretable regression extension for the Ex-Fuzzy library, enabling Mamdani fuzzy inference with scalar consequents learned directly from data. For this, a target-aware partition initialisation strategy based on Fuzzy C-Means clustering is introduced, in which linguistic variables are derived from an augmented input-output space to emphasise output-relevant regions of the feature space. The proposed extension is evaluated on ten regression datasets from the KEEL repository, comparing Gaussian and trapezoidal partition strategies against standard baselines including linear regression, multilayer perceptron, and random forests. Experimental results show that Gaussian partitions consistently outperform uniform trapezoidal partitions, achieving a mean coefficient of determination of approximately 0.86 while producing compact rule bases of 10-15 human-readable rules. The proposed implementation provides a transparent and competitive alternative to black-box regression models, supporting practical interpretability with competitive predictive performance.