Please use this identifier to cite or link to this item:
http://repository.ipb.ac.id/handle/123456789/125738
Title: | Optimasi jaringan syaraf tiruan menggunakan algoritme genetika |
Authors: | Herdiyeni, Yeni Rafdi, Muhammad Abi |
Issue Date: | 2010 |
Abstract: | Neural network is one of common computational model to recognize pattern. Backpropagation algorithm can be used as method to train the networks weights. One of the drawbacks of backpropagation training is the initial weights generated randomly. It causes the instability of neural network training performance. Because of this, many experiment and training must be conducted to obtain the desired model. This research combine genetic algorithm and neural network to overcome random generation problem. Genetic Algorithm (GA) is used to find the optimal initial weights based on learning using Mean Square Error as objective function. The initial weights using GA is expected to be stable and better than those generated. The experimental research showed that genetic algorithm can be used to optimize the performance of backpropagation neural network. The optimal initial weights from genetic algorithm is more stable than the initial weights from regular method. The classification accuracy of genetic algortihm and neural network is also higher than regular method. This indicated GA method is promising to increase the neural network performance and can be applied in various real problem. |
URI: | http://repository.ipb.ac.id/handle/123456789/125738 |
Appears in Collections: | UT - Computer Science |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
G10mar.pdf Restricted Access | Fulltext | 764.62 kB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.