Bacteria identification using artificial neural network: a case study of Peptococcaceae Family identification

Conventionally, bacteria identification is made through Bergey´s manual but this process is time consuming, requires full concentration and understandings and has the possibility for misidentifying of an unknown bacteria. In this study, artificial neural network was evaluated as a tool for identifying unknown bacteria. The study is conducted by collecting the data of microorganism properties from Bergey´s manual and then train the data by using neural network system in MATLAB programme. The study focused on facultative anaerobic bacterial species from Gram-Positive Cocci group under family Peptococcaceae. It is proposed that improved predictions can be obtained using a three-layer neural network instead of manually identifying the species using conventional and complicated method which is time-consuming and lengthy. The network training session was conducted by batch training using feed-forward back propagation algorithm. As a result, a neural network was successfully developed to accurately identify any bacterial species from Peptococcaceae family in a very short time.