THE IMPACT FACTORS ON THE ASSESSMENT OF SIMILARITY BETWEEN FUNCTIONS

Data processing, in any field, especially in the case of accessibility of relatively large volumes of data, becomes appropriate to involve more specific techniques, procedures, which at the initial stage could be considered as universal. An important factor in the development of algorithms for assessing the level of similarity between functions, in certain situations, depending on the nature of the phenomenon under research, may be the size of the variation interval of the independent variable. In this context, in this paper, certain suggestions, techniques for obtaining numerical characteristics, obtained based on methodologies for varying the lengths subintervals of the integral definition interval of approximating functions, deduced from the data set involved in the research, will be discussed. One of the main suggestions could be the division of the entire interval into several subintervals. The number of subintervals is supposed to be deduced depending on the nature of the phenomenon under investigation, thus assessing the level of similarity in each subinterval, and then building a synthesis algorithm for the integral interval. Such an algorithm – methodology – is to be presented in this paper, by presenting examples – case studies based on primary data similar to some real data. Another question that may arise is the nature of a possible real factor that could have a significant impact on the results of the similarity assessment. The essence of such factors may be very difficult to deduce from only a single data set. A solution would be to highlight the nature of several data sets related to the “circumstances” of data production in the research process, as well as to their collection methodologies.