Information Distribution
Chongfu Huang, Yong Shi · Studies in fuzziness and soft computing · 2002
This chapter presents the concept of information distribution. Through information distribution, we can change the crisp observations of a given sample into fuzzy sets. Hence, fuzzy sets are employed to describe the fuzzy transition information in a small sample. It is useful to improve the estimation of the probability distribution. Based on this estimation, we can construct fuzzy relationships, directly, without any assumptions. In detail, we discuss the method of 1-dimension linear-information-distribution. Computer simulation shows the work efficiency of the new method is about 23% higher than the histogram method for the estimation of a probability distribution. The chapter is organized as follows: in section 4.1, we introduce the concept of information distribution. In section 4.2, we give the mathematical definition of information distribution. Section 4.3 gives the method of 1-dimension linear-information-distribution. Section 4.4 demonstrates the benefit of information distribution for probability distribution estimation. In section 4.5, we construct a fuzzy relation matrixes with the method of information distribution. In section 4.6, we discuss approximate inference based on information distribution. These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.