Team: 30
School: La Cueva High
Area of Science: Mycology
Proposal: Mushroom hunting is a beloved hobby, especially in Europe. But despite the existence of numerous datasets, many apps still misdiagnose common poisonous mushrooms as safe. This problem has negatively impacted the prospects of mushroom-hunting, where hikers are further discouraged from picking mushrooms that are unfamiliar to them.
I plan to use multiple datasets (a dataset with mushroom characteristics, and a dataset with mushroom images) to classify mushrooms as edible or poisonous. The goal of my project is to take a picture of a mushroom, extract the relevant characteristics, and predict the edibility from those characteristics.
I have a dataset from the University of California Irvine with 8000 mushroom entries from the Agaricus and Lepiota family. Each entry contains a list of 22 mushroom characteristics and the corresponding edibility. Using machine learning functions in the Wolfram Language, I first plan to find the specific mushroom characteristic(s) that correlate the most to edibility. Afterwards, I plan to use mainly the Classify (characteristics) and FeatureExtract (images) functions to predict edibility given those characteristics.
Team Members:
Sponsoring Teacher: Yolanda Lozano