Identifying living organisms by using artificial neural networks approach / Raifiza Abdul Rahim

Abdul Rahim, Raifiza (2003) Identifying living organisms by using artificial neural networks approach / Raifiza Abdul Rahim. Student Project. Faculty of Information Technology & Quantitative Sciences, Shah Alam. (Unpublished)

[img]
Preview
Text
PPb_RAIFIZA ABDUL RAHIM CS 03_5.pdf

Download (108kB) | Preview

Abstract

Identifying DNA sequences is very useful in forensic area. Currently, there are a lot of computational biology approaches (bioinformatics) in solving the molecular biology. The variation, complexity, and incompletely-understood nature of sequences make it impractical to hand-code algorithm by applying the human ability and laboratory equipments in identifying the sequences. Artificial neural network (ANN), which is one of the commonly used machine learning technique, might be preferable to form its own descriptions of genetic concepts. Thus, it is applied in developing a prototype to identify the living organism whether it is human (Malay or India) or non-human. A multi-layer backpropagation algorithm of one hidden layer with 5 neurons was used. It is the constant representation whereby it produces one output. The training set was composed of 7 types of organisms from randomly selected DNA nucleotide sequences. The result of this prototype shows that it successfully can train the sequence of non-human. The reason it cannot train the human sequence probably because the way of massaging the data. By using the different sequences from the same types of organisms, the network successfully can identify. The training epoch and time can be accelerated if the network is included with the momentum.

Item Type: Monograph (Student Project)
Uncontrolled Keywords: neural networks, DNA, forensics, molecular biology
Subjects: Q Science > QA Mathematics > Instruments and machines > Electronic computers. Computer science > Neural networks (Computer science)
Q Science > QA Mathematics > Instruments and machines > Electronic computers. Computer science > Neural networks (Computer science)

Q Science > QH Natural history - Biology > DNA. Genetic code. Transcription
Divisions: Faculty of Information Technology and Quantitative Sciences
Depositing User: Staf Pendigitan 1
Date Deposited: 28 Dec 2010 01:49
Last Modified: 03 Jul 2017 06:33
URI: http://ir.uitm.edu.my/id/eprint/945

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year