Teknik Tapisan Dalam Pemodelan Perceptron Multi Aras

• Main Author: Roshidi, Din
• Format: Thesis
• Language: English; English
• Published: 1999
• Subjects: QA76 Computer software
• Online Access: http://etd.uum.edu.my/708/1/ROSHIDI_BIN_DIN_-_Teknik_Tapisan_Dalam_Pemodelan_Perceptron_Multi_Aras.pdf; http://etd.uum.edu.my/708/2/1.ROSHIDI_BIN_DIN_-_Teknik_Tapisan_Dalam_Pemodelan_Perceptron_Multi_Aras.pdf; http://etd.uum.edu.my/708/

This research is done to see the implication filtering technique on house price modeling performance in predicting based on pre-processing technique in learning capability in neural network using Multilayer Perceptron. This research also discuss the relationship between house price analysis and neural network application in predicting terrace house price. This research uses functional mathematics approach based on previous years house price index to predict the actual value of terrace house in the future. This approach is also used in predicting medium period by collecting actual terrace house price data in Kuala Lumpur. These data is not only based on terrace house price index but also based on various aspect which involve direct or indirectly. The obtained result is compared with actual price index the following year. Furthermore, the finding gained from the research model shows that is used will produce better prediction.