A New Approach of Disaster Forecasting Based on Least Square Optimized Neural Network

Table 1

Data simulation of different hidden layer neuron number.

Training data	Prediction results of different hidden layers					Actual results
	12	13	14	15	16
Name A	0.907	0.872	0.915	0.957	0.960	1
Name B	0.644	0.529	0.514	0.506	0.704	0
Name C	0.811	0.971	0.869	0.967	0.986	1
Name D	0.526	0.521	0.507	0.503	0.561	0
Name E	0.960	0.948	0.905	0.917	0.914	1
Name F	0.946	0.896	0.880	0.948	0.927	1
Name G	0.533	0.556	0.509	0.522	0.535	0
Name H	0.898	0.903	0.928	0.954	0.958	1
Name I	0.503	0.579	0.689	0.518	0.502	0
Name J	0.960	0.921	0.939	0.945	0.940	1