Hand Held Electronic Nose System Using Machine Learning and Predictive Modeling Techniques-A supervised classification-based approach

Electronic Nose is one of a number of technological models that can be harnessed to its full potential using Machine Learning. It has the ability to predict whether a specific food sample is rotten or not based on the gases released by the food sample. In this paper, we have discussed how we have been able to achieve accurate predictions with a specific food sample being spoilt or not by training the model on previously recorded data and testing it on unseen data using supervised classification models. By using a number of popular Machine Learning models like K-Nearest Neighbor, Support Vector Machine, Random Forest ensemble and Logistic Regression to train and predict the results, we have been able to predict the best and accurate result which will provide a real-life solution to an age-old problem of detecting whether a food sample is rotten or not. In this work, Python is used as the software tool.