Identification accuracy of sample distribution functions for time series depending on distribution type, norm and sample length
Abstract:
Statistical methods of sample distribution identification have been investigated for several types of general distributions – unimodal, bimodal and distributions with high dispersion. For various types of general distributions the optimal identification depends on norm and sample length. The optimal level of separation is determined as a result of statistical experiments.
Keywords:
Identification accuracy of sample distribution functions for time series depending on distribution type, norm and sample length
Publication language:russian, pages:25
Research direction:
Mathematical modelling in actual problems of science and technics
