An improved method for eliminating the outlier values in small sample

Hua Meng; Jie Li; Yang Liu

doi:10.1117/12.3029212

6 May 2024 An improved method for eliminating the outlier values in small sample

Hua Meng, Jie Li, Yang Liu

Proceedings Volume 13107, Fourth International Conference on Sensors and Information Technology (ICSI 2024); 1310745 (2024) https://doi.org/10.1117/12.3029212
Event: Fourth International Conference on Sensors and Information Technology (ICSI 2024), 2024, Xiamen, China

Abstract

Under the condition of the small sample^[1,2,3,4,5], it is no longer possible to simply use the mean and variance of all the data to remove outlier values. Method of hampel is simple and rough in small sample, it will be greatly affected by the outlier value that once enters the middle 50% of all the data. Method of boxplot^[6,7,8] will cause the interquartile range (IQR) to be too large and make the outlier value become normal. In order to solve the above problems, this paper improves the method of boxplot by using the median to eliminate the outlier value in the middle 50% of all the data. Then we use the remaining data to calculate the mean value and variance. At last by use of the 3δ principles of method of hampel, we eliminate the outlier values and obtain the true and non-outlier values. The effect of the algorithm of this paper is useful and ideal by data simulation. It can be applied to engineering practice.

(2024) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Hua Meng, Jie Li, and Yang Liu "An improved method for eliminating the outlier values in small sample", Proc. SPIE 13107, Fourth International Conference on Sensors and Information Technology (ICSI 2024), 1310745 (6 May 2024); https://doi.org/10.1117/12.3029212

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available