Artificial Neural Network can solve a problem based on its experience and training. The purpose of this research is analyzing the comparison of two training that is Backpropagation Training (BP) and Levenberg-Marquardt (LM) Training to determine the speed and accuracy of each training. The case that be used is the weather prediction at Makassar City which is obtained from BMKG. The advantage of BP is the runtime execution to get SSE’s value faster than LM that has difference 410.48974 ms. The advantage of LM is can obtain the smallest SSE’s value on training process than BP, that is 0.0264.
Copyrights © 2017