Prediction of Insufficient Accuracy for Mushroom Classification whether Poisonous or Eatable Mushrooms using Support Vector Machine by comparing Logistic Regression to Improve Accuracy

Abstract

Aim: The main objective of the research study is to improve the good quality production for the better accuracy of eatable or
poisonous mushrooms. It ought to be checked for consumable mushrooms. Exact assurance and appropriate distinguishing proof of
species are the main safe approach to guarantee that the eatable mushrooms are poisonous, and defend against potential mishaps of
burning through harmful one. Materials and Methods: The study used 51 samples with two groups of algorithms and the G-power worth of 80%. The mushrooms were collected from various web sources with recent study findings and threshold 0.05%, confidence interval 95% mean and standard deviation. To predict the Mushroom accuracy rate for already the Logistic Regression (LR) algorithm has found 90.46% of accuracy, therefore this study needs to find better accuracy for accuracy decrease prediction with the Novel Support Vector Machine (SVM) in Machine Learning. Results: This research study found 91.37% of accuracy for poisonous detection on food using the Support Vector Machine algorithm with a statistically significant difference between the two groups (p=0.022; p<0.05) with 95% confidence interval. Conclusion: This study concludes that the Novel Support Vector Machine algorithm on accuracy is significantly better than the Logistic Regression algorithm.