Title :
Maximum Likelihood Logistic Regression Using Metaheuristics
Author :
Peterson, Leif E.
Author_Institution :
Center for Biostat., Methodist Hosp. Res. Inst., Houston, TX, USA
Abstract :
Maximum likelihood-based logistic regression coefficients and fitness growth rates for several metaheuristic techniques were compared with results from Newton-Raphson iteration. Metaheuristics included genetic algorithms (GA), covariance matrix self-adaptation evolution strategies (CMSA-ES), particle swarm optimization (PSO), and ant colony optimization (ACO). Results indicate that fitness growth rates for GA were greatly inferior to fitness values for NR, CMSA-ES, PSO, and ACO. For the data sets considered, coefficients determined using CMSA-ES were identical to coefficients generated with NR, while coefficients from PSO- and ACO-based logistic regression were only slightly different. For the ionosphere data with a larger number of features, ACO likelihood fitness growth was slower when compared with CMSA-ES and PSO. Because this was an early investigation of metaheuristics in logistic regression, future studies employing similar metaheuristics should focus on investigation of global vs. local minima.
Keywords :
covariance matrices; genetic algorithms; maximum likelihood estimation; particle swarm optimisation; regression analysis; Newton-Raphson iteration; ant colony optimization; covariance matrix self-adaptation evolution strategies; fitness growth; genetic algorithms; global minima; ionosphere data; local minima; maximum likelihood logistic regression; metaheuristics; particle swarm optimization; Ant colony optimization; Cost function; Covariance matrix; Genetic algorithms; Ionosphere; Logistics; Machine learning; Optimization methods; Particle swarm optimization; Runtime;
Conference_Titel :
Machine Learning and Applications, 2009. ICMLA '09. International Conference on
Conference_Location :
Miami Beach, FL
Print_ISBN :
978-0-7695-3926-3
DOI :
10.1109/ICMLA.2009.140