Bacteria identification via Artificial Neural Network based-on Bergey’s manual

Due to limitations and disadvantages of current identification process of unknown bacteria, artificial neural network can be employed as an alternative technique in bacteria identification at low cost and less time consuming. Artificial Neural Network (ANN) is a developed biological neurons principle based system in MATrix LABoratory (MATLAB) computer software that connects known input value to desired output or target value of the system and with the help of Bergey’s manual as data sources of chemical and physical characteristic of selected bacteria. Therefore, unknown bacteria can be successfully determined. Analyzing and data extraction from Bergey’s manual require high understanding of selected microorganisms in order to prevent any error or inaccurate result generated from ANN. Therefore, this study was conducted on Gram-Negative Bacillus shape bacteria under Betaproteobacteria Class and Order Hydrogenophilales. Selected bacteria under Hydrogenophilales order was Bacteria family of Hydrogenophilaceae. Levenberg Marquardt algorithm based Feed-forward backpropagation with Multilayer perceptron type of ANN was used in the training and learning sessions of the ANN development in order to obtain high accuracy simulation results within short period of time.