Report for Q3 : MLP Regressor

Created on October 23|Last edited on November 20
Comment
﻿
Hyperparameter Tuning﻿
Training R2
Training R2
Showing first 10 runs
0100200300400Step-5-4-3-2-10
Validation R2
Validation R2
Showing first 10 runs
0100200300400Step-6-4-20
Validation MSE
Validation MSE
Showing first 10 runs
0100200300400Step2468
Training MSE
Training MSE
Showing first 10 runs
0100200300400Step123456
Run set15
﻿
Best hyperparameters for both the training and the validation set are : sgd-tanh-0.001
﻿
Hyperparameters			Validation MSE 				Training MSE
batch-relu-0.0005			0.5532149593017942			0.7608131641366255
batch-tanh-0.0005			0.4815025720601651			0.7100752060167854
mini-batch-tanh-0.0005		0.3238408234191069			0.3510402434745661
mini-batch-tanh-0.0005		0.23969210195292326		0.34096707154637645
batch-tanh-0.1				0.16511484545759936		0.11844068269299524
batch-relu-0.1				0.10284189547005891		0.09102103171508302
sgd-sigmoid-0.1				0.09696675032274003		0.04332080687699516	
sgd-tanh-0.001				0.07729714106467031		0.05169007520861512
﻿
﻿
﻿
Add a comment