Research Article
A New Model for Building Energy Modeling and Management Using Predictive Analytics: Partitioned Hierarchical Multitask Regression (PHMR)
Table 3
MSPE and Pearson’s correlation on testing data for each leaf node (mean (std) over 100 simulation runs) size.
| | MSPE | Pearson’s correlation () | Sample size: 2000 | Sample size: 1000 | Sample size: 2000 | Sample size: 1000 |
| Node 2 | 0.25 (0.01) | 0.26 (0.02) | 1.00 (0.00) | 1.00 (0.00) | Node 4 | 0.26 (0.02) | 0.26 (0.02) | 1.00 (0.00) | 1.00 (0.00) | Node 6 | 0.26 (0.03) | 0.30 (0.24) | 1.00 (0.00) | 1.00 (0.00) | Node 8 | 2.08 (2.86) | 5.00 (3.94) | 0.99 (0.02) | 0.96 (0.03) | Node 9 | 2.18 (2.76) | 5.60 (3.92) | 0.99 (0.02) | 0.96 (0.03) |
|
|