Outlier Detection Method in Linear Regression Based on Sum of Arithmetic Progression

Table 4

EMMS for identifying an outlier.

𝑋(𝑛−1)	𝑦(𝑎𝑛)	𝑦^𝑇(𝑎𝑛−𝑎1)	𝑦^𝑇𝑇(|𝑦^𝑇−𝑥𝑛(𝐺^𝑇𝑦/𝐺𝑥)|)
0	100	0.000	0
1	101	1.000	0.06
2	102	2.000	0.12
3	103.6	3.600	0.42
4	104	4.000	0.24
𝐺𝑥=2		𝐺^𝑇𝑦=2.120
		EMMS (Max)	0.500
		EMMS (Min)	0.333
		2/𝑛	0.4
		Outlier?	Yes-Max