A Neuro-Fuzzy Model for Function Point Calibration

TL;DR

This paper presents a neuro-fuzzy model to calibrate Function Point weights, improving software effort estimation accuracy by 22% through empirical validation with industry data.

Contribution

It introduces a novel neuro-fuzzy approach for Function Point calibration, combining neural learning and fuzzy logic to enhance effort estimation accuracy.

Findings

22% improvement in effort estimation accuracy

Empirical validation with ISBSG data

Effective calibration of FP complexity weights

Abstract

The need to update the calibration of Function Point (FP) complexity weights is discussed, whose aims are to fit specific software application, to reflect software industry trend, and to improve cost estimation. Neuro-Fuzzy is a technique that incorporates the learning ability from neural network and the ability to capture human knowledge from fuzzy logic. The empirical validation using ISBSG data repository Release 8 shows a 22% improvement in software effort estimation after calibration using Neuro-Fuzzy technique.