Abstract. Tuning hyper-parameters is a necessary step to improve learning algorithm performances. For Support Vector Machine classifiers, adjusting kernel parameters increases dra...
Originally, genetic algorithms were developed based on the binary representation of candidate solutions in which each conjectured solution is a fixed-length string of binary numb...
We present a Semantic Optimized Service Discovery (SemOSD) approach capable of handling Web service search requests on a fine-grained level of detail where we augment semantic ser...
Many classification algorithms use the concept of distance or similarity between patterns. Previous work has shown that it is advantageous to optimize general Euclidean distances (...
— The chaos game, in which a moving point is repeatedly averaged toward randomly selected vertices of a triangle, is one method of generating the fractal called the Sierpinski tr...