Optimization of fuzzy analogy in software cost estimation using linguistic variables

TL;DR

This paper introduces a fuzzy logic-based optimization approach for software cost estimation that effectively handles categorical and linguistic project attributes, improving prediction accuracy and interpretability.

Contribution

The paper proposes a novel fuzzy analogy method utilizing linguistic variables and quantifiers to better manage imprecision in software effort estimation.

Findings

Improved estimation accuracy on NASA dataset.

Enhanced interpretability of cost models.

Outperforms traditional machine learning methods.

Abstract

One of the most important objectives of software engineering community has been the increase of useful models that beneficially explain the development of life cycle and precisely calculate the effort of software cost estimation. In analogy concept, there is deficiency in handling the datasets containing categorical variables though there are innumerable methods to estimate the cost. Due to the nature of software engineering domain, generally project attributes are often measured in terms of linguistic values such as very low, low, high and very high. The imprecise nature of such value represents the uncertainty and vagueness in their elucidation. However, there is no efficient method that can directly deal with the categorical variables and tolerate such imprecision and uncertainty without taking the classical intervals and numeric value approaches. In this paper, a new approach for optimization based on fuzzy logic, linguistic quantifiers and analogy based reasoning is proposed to improve the performance of the effort in software project when they are described in either numerical or categorical data. The performance of this proposed method exemplifies a pragmatic validation based on the historical NASA dataset. The results were analyzed using the prediction criterion and indicates that the proposed method can produce more explainable results than other machine learning methods.