Title: DESA: Differential Evolution-Simulated Combinations of Gaussians to Data Annealing for Fitting Linear Author: Rizavel Corsino-Addawe Degree: Master of Science in Statistics Abstract: The method of Differential Evolution (DE) is applied to the problem of fitting linear combinations of Gaussians to data. The DE method offers advantages over competing methods such as the Simulated Annealing (SA) algorithm. Firstly, the DE does not require an auxiliary special method to obtain initial parameter estimates as in SA algorithm. Secondly, based on experimental computer runs, the DE method is faster and gives more accurate results than those obtaines by SA. This paper also illustrates the combination of the search capabilities of DE and SA to develop a hybrid algorithm, DESA, which is equally applicable and has better searching ability and power to reach a near optimal solution.