Advances in Artificial Neural Systems

Research Article

Comparison of Artificial Neural Network Architecture in Solving Ordinary Differential Equations

Table 2

Comparison of the results (Example 1).
Input dataAnalytical Euler Runge-Kutta   
(Six nodes)
01.00001.00001.00001.0000
0.05000.95360.95000.95360.9677
0.10000.91370.90720.91380.9159
0.15000.87980.87070.87990.8815
0.20000.85140.84010.85150.8531
0.25000.82830.81500.82830.8264
0.30000.81040.79530.81050.8114
0.35000.79780.78100.79790.7953
0.40000.79050.77210.79070.7894
0.45000.78890.76890.78900.7845
0.50000.79310.77170.79320.7957
0.55000.80330.78050.80350.8041
0.60000.82000.79580.82010.8204
0.65000.84310.81780.84330.8399
0.70000.87310.84670.87330.8711
0.75000.91010.88260.91020.9151
0.80000.95410.92580.95420.9555
0.85001.00530.97631.00540.9948
0.90001.06371.03421.06381.0662
0.95001.12931.09951.12941.1306
1.0001.20221.17211.20221.2058