Complexity Weights Parameter Optimization of Use Case Points Estimation using Chaotic PSO

Use Case Point has been used in software effort estimation to calculate a project cost. One of the parameters used is the Use Case complexity weight which significantly affects estimation accuracy. Nevertheless, the current complexity weight levels result in unreliable measurement and abrupt classification caused by discontinuous weight levels. This paper proposes a chaotic Particle Swarm Optimization for optimizing the Use Case complexity weight parameters. The optimum complexity weight parameters that minimize the mean absolute error are selected using the particle swarm optimizer. The proposed algorithm is compared with standard PSO to evaluate their performance. The experimental procedure showed that Bernoulli chaotic map yielded the best mean solution of 1008.82 and was statistically significant with
$p$
-values less than 0.05 (0.018). The experiment proved that the proposed algorithm is robust in finding the optimal Use Case complexity weight.