Improving The Classification Of Banana Maturity Level With Hybrid Committee Of Ann

Fadhilah, Sapian (2008) Improving The Classification Of Banana Maturity Level With Hybrid Committee Of Ann. Project Report. UTeM, Melaka,Malaysia. (Submitted)

	PDF (24 Pages) Improving_The_Classification_Of_Banana_Maturity_Level_With_Hybrid_Committee_Of_Ann_FadhilahBinti_Sapian_QA76.87.F32_2008_-_24_Pages.pdf - Submitted Version Download (239kB)
	PDF (Full Text) Improving_The_Classification_Of_Banana_Maturity_Level_With_Hybrid_Committee_Of_Ann_FadhilahBinti_Sapian_QA76.87.F32_2008.pdf - Submitted Version Restricted to Registered users only Download (1MB)

Abstract

This project is to predict and improve the classification of banana maturity level using hybrid committee of Artificial Neural Network (ANN). The main objective of this project is to improve the accuracy of classification in identifying banana maturity level as an alternative way rather than press the banana. This accuracy depends on several ANN parameters such as number of hidden layers and nodes, learning rate,momentum rate and training function. The back propagation which consist of multi-layered perceptron make it possible to train the ANN training patterns. An array of seven neurons corresponds to the four indicator that related to each banana maturity‟s level, average, standard deviation and variance of banana maturity‟s level are the input for training phase. After passing through testing phase with the optimum training network, hybrid committee of two different networks give the best result in identifying then improving the classification of banana maturity‟s level.

Actions (login required)

View Item