Escherichia coli growth modeling using neural network

Abstract

The assessment of water microbial quality is normally performed by verification of
Escherichia coli where the growth is in nonlinearity. NARX is computational tools that have
extensive utilization in solving nonlinear time series problems. It is well known as one of the
technique that has the ability to predict with efficient and good performance. Using NARX, a
highly accurate model was developed to predict the growth of Escherichia coli (E. coli) based
on pH water parameter. The multiparameter portable sensor and spectrophotometer data were
used to build and train the neural network. The selection of neural network structure for pH
and optical density modelling was optimized and also the training and validation were
analyzed. The result exhibited that NARX modelling was able to predict the growth of E. coli
based on pH water parameter with overall regression is 0.99956.

Keywords: neural network; NARX; prediction; Escherichia coli; pH; optical density.