The Importance of Big Data Preparation in Microbial Colony Classification

Microbe is a microscopic organism that has been used in various industrial and medical processes. Unfortunately, microorganisms can also cause various diseases that are detrimental to humans. It is estimated that there are around 1400 types of pathogens that have the potential to induce disease in humans. Microbial classification is needed to determine whether a microorganism is beneficial or harmful to humans. After we researched the importance of data preparation and a CNN Model, InceptionV3, our results have reached accuracy of 0.94 with Macro average precision of 0.95, recall 0.96, and f1-score 0.95. We have also reached weighted average precision of 0.95, recall 0.94, and f1-score 0.94. In conclusion, we gained proves that even with the small number of images which ranges from various widths, shapes, colors, and sizes, can still provide correct classification results to a 89% accuracy.