BACKPROPAGATION ARTIFICIAL NEURAL NETWORK TO IDENTIFY FOOTBALL PLAYER OPTIMAL POSITION
Artificial neural network uses the processing of the brain as a basis to develop algorithms that can be used to model complex patterns and prediction problems. The capability of this model on recognizing pattern of the data has lead this model to be useful for data classification. Many method has be...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/28383 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | Artificial neural network uses the processing of the brain as a basis to develop algorithms that can be used to model complex patterns and prediction problems. The capability of this model on recognizing pattern of the data has lead this model to be useful for data classification. Many method has been developed in artificial neural network. The most common one is backpropagation. In this final project, we focus on using backpropagation artificial neural network for classifiying football player data. The data consists of 6 technical attributes namely pace, shooting, passing, dribling, defense, and physic; to find optimal position of football player as defender, midfielder, or attacker. The available data is divided into two groups : training and testing data. Some neural network architectures are conducted using R-programming language. The result shows that the neural network architecture with one input layer (6 neurons), three hidden layer (each layer has 6 neurons), and one output layer (3 neurons) gives the best accuracy. The architecture use sigmoid biner as activation function of each neuron, sum-of-squares error function, and batch gradient-descent optimization for each training. |
|---|