A Least Squares Method for Variance Estimation in Heteroscedastic Nonparametric Regression

Table 2

The relative MSEs of 𝜎²₁, 𝜎²₂, and 𝜎²_TW for scenario (ii) under various simulation settings.

𝑛	𝜎	𝑓	𝜎^21	𝜎^22	𝜎^2TW
30	0.5	𝑓1	5.356385	5.358625	5.76456
		𝑓2	5.825295	5.844773	6.410941
		𝑓3	24.08339	24.74582	27.14087
	2	𝑓1	5.291982	5.292094	5.659067
		𝑓2	5.284907	5.285037	5.659967
		𝑓3	5.367277	5.370764	5.809319
100	0.5	𝑓1	14.70915	14.70799	15.1483
		𝑓2	14.6843	14.68265	15.1374
		𝑓3	15.01971	15.02443	15.58368
	2	𝑓1	14.71244	14.71141	15.14792
		𝑓2	14.70282	14.70165	15.13869
		𝑓3	14.69388	14.6925	15.13526
500	0.5	𝑓1	42.08678	42.08626	42.35964
		𝑓2	42.09082	42.0904	42.36408
		𝑓3	42.10065	42.10021	42.37591
	2	𝑓1	42.0839	42.08347	42.3567
		𝑓2	42.08482	42.08442	42.35764
		𝑓3	42.08672	42.08632	42.35965