摘要：Optimization methods commonly aredesigned for solving the optimization problems. Local search algorithms are optimizationmethod, which are good candidate in exploiting the search space. However, most ofthem need parameter tuning and incapable of escaping from local optima. This workproposes non-parametric Acceptance Criterion (AC) that not relies on user-defined,which motivate to propose an Adaptive Acceptance Criterion (AAC). AC accepts a littleworse solution based on comparing the candidate and best solutions found valuesto a stored value. The value is stored based on the lowest value of comparing thecandidate and best solution found, when a new best solution found. AAC adaptivelyescape from local optima by employing a similar diversification idea of a previousproposed (ARDA) algorithm. In AAC, an estimated value added to the threshold (whenthe search is idle) to increase the search exploration. The estimated value is generatedbased on the frequency of the solutions quality differences, which are stored inan array. The progress of the search diversity is governed by the stored value.Six medical benchmark datasets for clustering problem (which are available in UCIMachine Learning Repository) and eleven benchmark datasets for university coursetimetabling problems (Socha benchmark datasets) are used as test domains. In orderto evaluate the effectiveness of the propose AAC, comparison made between AC, AACand other approaches drawn from the scientific literature. Results indicate that,AAC algorithm is able to produce good quality solutions which are comparable toother approaches in the literature.
关键词：Course Timetabling Problem ; Multi K-Means; Medical Clustering Problems; Adaptive Acceptance Criterion; Local Search Algorithms
