Execution Time Modeling on FastDeepIoT Profiling Dataset Nexus 5 (test)
1.8MAPEFastDeepIoT
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| FastDeepIoTHardware=Nexus 5, Layer Type=Gated Recurrent Unit, Input Feature=Explanatory variable vector x2018.09 | 1.8 | 0.6 | 0.999 | |
| DNNHardware=Nexus 5, Layer Type=Gated Recurrent Unit, Input Feature=Explanatory variable vector x, Architecture=4-layer fully connected2018.09 | 2 | 0.7 | 0.999 | |
| FastDeepIoTDevice=Nexus 5, Target Layer=Long short term memory, Input Feature=Explanatory variable vector x2018.09 | 2.3 | 0.6 | 0.999 | |
| DNNDevice=Nexus 5, Target Layer=Long short term memory, Input Feature=Explanatory variable vector x2018.09 | 2.8 | 0.9 | 0.998 | |
| GBRTDevice=Nexus 5, Target Layer=Long short term memory, Input Feature=Explanatory variable vector x2018.09 | 4.1 | 1.6 | 0.993 | |
| RFDevice=Nexus 5, Target Layer=Long short term memory, Input Feature=Explanatory variable vector x2018.09 | 4.7 | 1.4 | 0.995 | |
| GBRTHardware=Nexus 5, Layer Type=Gated Recurrent Unit, Input Feature=Explanatory variable vector x2018.09 | 4.8 | 1.5 | 0.995 | |
| RFHardware=Nexus 5, Layer Type=Gated Recurrent Unit, Input Feature=Explanatory variable vector x2018.09 | 6.7 | 1.8 | 0.995 | |
| DTDevice=Nexus 5, Target Layer=Long short term memory, Input Feature=Explanatory variable vector x2018.09 | 9 | 3 | 0.977 | |
| DTHardware=Nexus 5, Layer Type=Gated Recurrent Unit, Input Feature=Explanatory variable vector x2018.09 | 9.4 | 2.9 | 0.986 | |
| SVRDevice=Nexus 5, Target Layer=Long short term memory, Input Feature=Explanatory variable vector x2018.09 | 73.7 | 23.7 | -0.223 | |
| SVRHardware=Nexus 5, Layer Type=Gated Recurrent Unit, Input Feature=Explanatory variable vector x, Kernel=Radial basis function2018.09 | 78.7 | 23.6 | -0.078 |