Mushroom Classifier System Using Machine Learning Algorithm

Abstract

This paper studies the classification method for mushrooms based on Machine Learning algorithms and Graphic Processors. As mushrooms have many species, classification of mushrooms turns out to be difficult. These large numbers of species contain some edible and some poisonous or deadly poisonous mushrooms. Sometimes mushroom recognition is difficult through our naked eyes and due to a lack of knowledge of the identification of edible mushrooms, recognition turns out to be complicated. Although there are experts in distinguishing poisonous mushrooms from the list of edible mushrooms, where 70-80 species are reported as poisonous, occasional cases occur of misidentification of fatal mushrooms. Also, mushroom collectors have no formal discipline for a testimony of mushrooms, due to which people consume such wild mushrooms misidentified as nutritious mushrooms resulting in life-threatening Categorical features, support vector machine, convolution neural network, graphics processor, machine learning g disease or death-causing illnesses. The main aim of this project was to apply a method to detect whether the mushrooms fit human consumption or not. This paper proposes a method to determine the quality of the mushrooms using a categorical dataset that has 23 distinct characteristics. To solve the complication of the classification of mushrooms, a supervised learning model with the associated learning model is used. This method achieves a good result through the comparison of total time and speed between GPU and CPU.