Classification of E.coli, Lactobacillus and Bacillus Subtilis Using Computational Intelligence Approach

Abstract

Classification of bacterial species is an important thing in the biochemical sciences. The right distinguishing proof of microorganisms is of essential significance to microbial systematists and in addition to researchers required in numerous different zones of connected research and industry. Practically, it can be time consuming and costly as well.

The work in this thesis mainly focuses on prediction of bacterial classes using machine learning methods. Its objective is to find out the optimum parameters for the bacterial classification from the measurable features such as concentration, absorbance and pH values of the given solution.

We collected the practically performed data and arranged it in order for simulating it in R. Four different bacterial species were taken and their ordered data was simulated in R. Four machine learning models were used i.e. Random forest, Decision tree, SVM and Linear model with eight different parameters.

Comparison of the performances of each of the applied machine learning model was done to know about the most accurate model. And at last, k fold cross validation was done in order to investigate the robustness of the best fit model.