An Intelligent System for Eliminating the Suspicious Experiment Data
Jiang Xingfang1, Chen Dongdong2, Fu Haiming2, Huang Shuang2, Liu Hongfei3
Citation : Jiang Xingfang,et.al, An Intelligent System for Eliminating the Suspicious Experiment Data International Journal of Research Studies in Computer Science and Engineering 2014, 1(3) : 15-20
There is a distinct suspicious experiment datum (outlier) in a group data and they could be eliminated by the method of three times standard deviation. For the problem the method of two times standard deviation has been proposed and the intelligent system for eliminating the outlier had been developed. The intelligent system has four stages. The first one is that the general calculation software has been developed for calculating the mean and the uncertainity in the ten data. The second one is that the calculation software is for filling data randomly by user. The thrid one is suit for filling data randomly by user and there are outliers in the experiment data. The fourth one is that the intelligent system is suit for filling randomly by user, there are outliers, and the minimum digit of decimal places being different in experimental data. The statistics method of "add 1 for not empty" for filling randomly by user has been adopted. The confidence possibility of 95.5% that it is the judgment method of two times standard deviation range near the mean has been proposed. The method of "reading characters" for calculating the minimum digit of decimal places has been applied. The intelligent system is good for all the data processing in Physics experiment and it could be spreaded to other subjects for data processing