Member-only story
Leveraging Feature Importance to Predict Mushrooms’ Edibility in Python
Leveraging Feature Importance to predict whether mushroom species are edible or poisonous
Introduction
This article aims at leveraging feature importance to assess whether all the columns within a dataset need to be used for prediction or not.
Fall’s approaching. Imagine you’re enjoying a walk in the woods and you find some mushrooms on the side of the path. Wouldn’t it be nice to input some of their features into an ML-powered application that can detect with confidence edible qualities? I’m personally not into mushroom hunting but I’m definitely into food, and I can already smell a nice dish of “tagliolini ai funghi” in front of me after a long walk.
Mushrooms are fungi, part of a kingdom of their own separate from plants and animals. As you might imagine, they present numerous features to assess before deciding whether one is edible or not.
Your trekking group is walking away and we need to take a decision as quickly as possible.
Which information is fundamental to comfortably assess a mushroom species’ edibility?
