Given their abundance in vital elements, proteins, minerals, and vitamins, it is advisable that we include mushrooms in our diet. Many farmers grow mushrooms in greenhouses, where the climatic conditions are specific and the space is constrained. Among the millions of varieties of mushrooms that exist on earth, there are two distinct types: edible mushrooms and harmful mushrooms. Many people get food poisoning because they are not aware that the mushrooms are poisonous. Even in some other countries, cases of poisoning from poisonous mushrooms have been documented. The difference between harmful and edible mushrooms might be challenging to make because of their similar characteristics and abundance. By assisting in the discovery of significant patterns from millions or even billions of data records, data mining techniques, specifically classification, can be used to determine the type of mushrooms. This paper offers a comparative investigation of different classification techniques for recognizing a dataset of mushrooms. Current research on classifying mushrooms focuses on using ML techniques individually, and some systems outperform others in terms of accuracy. Research on the classification of mushrooms is scarce. Based on this research, an integrated model was proposed that would combine rather than treat separately the decisions generated by the most accurate methodologies. The mushroom dataset was downloaded from the UC Irvine repository. The results show that the accuracy performance of the integrated model is 95% better than that of other techniques.

Machine learning, classification algorithm, edible mushroom, inedible mushroom, decision tree, Support vector machine, KNN etc