2015/09/22 by Malte Probst, Franz Rothlauf, Probst, Malte +1
Computer Science · #Advanced Neural Network Applications #FOS: Computer and information sciences #Generative Adversarial Networks and Image Synthesis #Human Pose and Action Recognition #Music and Audio Processing #Neural and Evolutionary Computing (cs.NE)
paper · pdf · doi:10.48550/arxiv.1509.06535
openalex publication_date 2015/09/22 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Estimation of Distribution Algorithms (EDAs) require flexible probability\nmodels that can be efficiently learned and sampled. Deep Boltzmann Machines\n(DBMs) are generative neural networks with these desired properties. We\nintegrate a DBM into an EDA and evaluate the performance of this system in\nsolving combinatorial optimization problems with a single objective. We compare\nthe results to the Bayesian Optimization Algorithm. The performance of DBM-EDA\nwas superior to BOA for difficult additively decomposable functions, i.e.,\nconcatenated deceptive traps of higher order. For most other benchmark\nproblems, DBM-EDA cannot clearly outperform BOA, or other neural network-based\nEDAs. In particular, it often yields optimal solutions for a subset of the runs\n(with fewer evaluations than BOA), but is unable to provide reliable\nconvergence to the global optimum competitively. At the same time, the model\nbuilding process is computationally more expensive than that of other EDAs\nusing probabilistic models from the neural network family, such as DAE-EDA.\n