PEMODELAN RANG KAI AN SUAP BALIK ELMAN BAGI PERAMALAN HARGA RUMAH

This paper discusses an Elman Recurrent Netivork for terrace house price prediction in Kuala Lumpur. The performance of the netivork with backpropagation algorithm is experimented and analysed to capture the behaviour of the housing data for better prediction. The prices of the terrace houses are influenced by eight factors and these factors become the input parameters for the network. The results show that the learning performance of the network implemented on the 80% of training data gives a better classification rate with an accuracy of 97.6% and an error value of 0.012744 on 20% of the 1997 test data.