Table 1.

GMM model fit summary with BIC values for inter- and intra-dwelling type.

Sl. no	Inter-dwelling type
	Model name	Best model	Optimal cluster (G)	Bayesian Information Criteria (BIC)	Log likelihood	Sample (n) [NA values are omitted]
1	2018	VEE	9	−16701.19	−8074.27	1439
2	2019	VEV	7	−15634.19	−7373.47	1440
3	2020	VVE	8	−4031.27	−1693.30	770
	Intra-dwelling type
4	BR1_2018	V	3	−1390.21	−668.52	770
5	BR1_2019	V	2	−1081.23	−511.52	1440
6	BR1_2020	V	4	−905.87	−2151.75	770
7	BR2_2018	V	2	−1630.86	−775.43	1439
8	BR2_2019	E	5	−189.53	−116.58	1440
9	BR2_2020	V	6	−612.41	−269.65	770
10	BR3_2018	V	3	−1558.88	−761.26	1439
11	BR3_2019	V	2	−2002.59	−927.20	1440
12	BR3_2020	V	2	−1381.77	−654.30	770
13	M3BR_2018	V	2	−713.10	−338.37	1439
14	M3BR_2019	V	3	−1053.66	−497.74	1440
15	M3BR_2020	V	6	−1925.94	1039.40	770
16	RM1_2018	V	3	−2582.27	−3468.70	1439
17	RM1_2019	E	2	−2053.39	−995.81	1440
18	RM1_2020	V	6	40.91	47.04	770

[Note: Optimal cluster (G) denote the value that fitted the GMM model value with lowest BIC values. Optimal G-values is in the range of 2–8 for inter-building clustering and a range of 2–14 for intra-building clustering. Log-likelihood values is used to validate the BIC-driven model fit. It is a function of sample size (n), and a higher value determines better fit (see section 3.3 for detail)].