College of Mechatronical Engineering, Henan University of Science and Technology, Luoyang, China

---

Abstract—Point estimation is an important issue in data processing. Classical statistics is taken into account to assess mainly the true value of a population under the condition of large sample sizes and known probability distributions. If sample sizes are small and probability distributions are unknown, many statistical methods may become ineffective. For this end, a fusion method based on the information poor theory is proposed. Synthetically considering the characteristics of the various methods and effectively depicting the true value from different aspects, the fusion method is able to obtain the final estimated true value used as the most appropriate representative of the true value of a population. The results of the Monte Carlo simulation and the experimental investigation indicate that the fusion method allows the number of the data to be little and the probability distribution to be unknown.

Index Terms—information fusion, point estimation, data series, small sample, information poor system

[PDF]

Cite: Xintao Xia, Leilei Gao, and Jianfeng Chen, "Fusion Method for True Value Estimation Based on Information Poor Theory," Journal of Software vol. 7, no. 5, pp. 1014-1021, 2012.