Application of hill climbing algorithm in determining the characteristic objects preferences based on the reference set of alternatives

Abstract

Random processes are a frequent issue when trying to solve problems in various areas. The randomness factor makes it difficult to clearly define the input parameters of a system in maximizing its effects. The solution to this problem may be the usage of stochastic optimization methods. In the following article, the Hill Climbing method has been used to solve the problem of optimization, which in combination with the COMET method gave satisfactory results by determining the relationship between the preference assessment of already existing alternatives to the newly determined alternatives. The motivation to conduct the study was the desire to systematize knowledge on the effective selection of input parameters for stochastic optimization methods. The proposed solution indicates how to select the grid size in an unknown problem and the step size in the Hill Climbing method.