Comparing Logistic Regression Algorithm with K-Nearest Neighbor Algorithm to Extract Features from Images for Prescriptive Analysis of Mushroom Contamination

In this project, we aimed to assess mushroom contamination by analyzing images using two different algorithms: a novel K-Nearest Neighbour algorithm and a traditional Logistic Regression algorithm. We tested both algorithms with a sample size of 45 images each, using Jupyter Notebook for the analysis. Our goal was to determine which algorithm better identifies toxic mushrooms. To measure how well each algorithm performed, we calculated the accuracy of both methods. We used SPSS software to analyze both the training and test datasets. The K-Nearest Neighbour algorithm outperformed the Logistic Regression algorithm in terms of accuracy, showing significantly better results. The statistical tests confirmed that the KNN algorithm's performance was notably superior, with a p-value less than 0.05, indicating that the difference in accuracy was statistically significant.