Predicts whether a mushroom is edible or toxic using classification models.

For a more in depth analysis, review the Jupyter Notebook.

Using the learn module we can visualize the performance of different classification models.

For example, we can use train_percent_accuracy to plot the relationship between each model's dataset training percentage and its accuracy score.

The lower accuracy at lower training percentages can be explained by not utilizing enough training data to effectively train each model.

As can be seen, Random Forest and Decision Tree classification were the most accurate, reaching scores of 1.00.

We can use feature_importances to visualize the importance of each feature when trained with the Random Forest Classifier:

The performance of each model can also be quantified. MultiClassifier.train_all populates the dictionary MultiClassifier.performances with the total time taken to train each model, which can then be plotted using performance_all: