Halton Based Initial Distribution in Artificial Bee Colony Algorithm and Its Application in Software Effort Estimation

Artificial Bee Colony (ABC) algorithm is an optimization algorithm based on the intelligent behaviour of honey bee swarm. ABC can be initialized with either a uniform or a non-uniform distribution. The decision regarding which to use depends on how much is known about the location of the optimum. Generally uniform distributions are preferred since they best reflect the lack of knowledge about the optimum's location. In this paper we have used Halton points for the initial distribution and compared the simulation results with rand (0,1) uniform distribution. Further the algorithm is a used to estimate the cost model parameters and than the performance is compared with the measured efforts on the NASA Software project dataset