Edge Impulse Inc. / Tutorial: temperature regression Public
Set the 'expected outcome' for each sample to the desired outcome to automatically score the impulse. Maximum absolute regression error is 5.9, set thresholds.
Sample name Expected outcome Length Accuracy Result Error
freezertemprampup.t=3.json.1u4cd6b5 3 2s 0% 21.93 18.93 (631.1%)
freezertemprampup.t=2.json.1u4cd6b4 2 2s 100% 6.96 4.96 (247.8%)
freezertemprampup.t=4.json.1u4cd6b5 4 2s 100% 2.33 1.67 (41.7%)
freezertemprampup.t=1.json.1u4c66r0 1 2s 100% 3.38 2.38 (238.1%)
freezertemprampup.t=40.json.1u4c66r0 40 2s 100% 37.8 2.20 (5.5%)
freezertemprampup.t=10.json.1u4c66r0 10 2s 100% 8.12 1.88 (18.8%)
freezertemprampup.t=30.json.1u4c66r0 30 2s 100% 26.25 3.75 (12.5%)
freezertemprampup.t=20.json.1u4c66r1 20 2s 100% 15.45 4.55 (22.8%)
freezertemprampup.t=50.json.1u4c66r1 50 2s 100% 48.66 1.34 (2.7%)
freezertemprampup.t=60.json.1u4c66qv 60 2s 100% 59.3 0.70 (1.2%)
freezertemprampup.t=5.json.1u4c66r1 5 2s 100% 8.44 3.44 (68.7%)
freezertemprampup.aftermiddle.json.1u47n0ll 18 2s 100% 13.56 4.44 (24.7%)
freezertemprampup.beforemiddle.json.1u47n0ll 40 2s 100% 37.37 2.63 (6.6%)
freezertemprampup.justbeforetheend.json.1u47c6j1 5 2s 100% 8.24 3.24 (64.8%)
freezertemprampup.beginning.json.1u471aja 60 2s 100% 59.3 0.70 (1.2%)
freezertemprampup.justafterbeginning.json.1u471ajd 58 2s 100% 56 2.00 (3.4%)
freezertemprampup.end.json.1u46e6n1 1 2s 100% 3.38 2.38 (238.1%)

Model testing output

Results

Model version:
Mean squared error 29.18
Mean absolute error 3.60
Explained variance score 0.94
Loading...
Loading...