Downscaling and Projection of Multi-CMIP5 Precipitation Using Machine Learning Methods in the Upper Han River Basin

Table 8

Optimal results of Bayesian hyperparameter optimization for MLP for region mean.

	Activation	Alpha	HLZ	LR	Max_iter	Solver	Tol	Objective
AC10	tanh	1.024201	20	Adaptive	1500	adam	0.00319	1444
AC13	logistic	9.831136	12	Adaptive	125	lbfgs	0.008452	1605
GG	tanh	4.26283	16	Adaptive	1152	adam	0.002589	1501
GH	tanh	8.535319	14	Adaptive	1385	adam	0.000246	1511
GM	logistic	3.838534	14	Adaptive	413	adam	0.006199	1528
GR	tanh	9.454777	12	Adaptive	1521	adam	0.003523	1755
IN	logistic	7.885871	11	Invscaling	145	lbfgs	0.005522	1520
Nor	logistic	9.299373	10	Constant	1511	lbfgs	0.000955	1458
MME	tanh	3.534963	6	Adaptive	834	adam	0.000142	1340
BMA	logistic	9.045651	9	Constant	1360	lbfgs	0.007592	1278

HLZ, hidden_layer_sizes; LR, learning_rate.