Shrooming - Interactive mushroom edibility predictions with XGBoost

Your task is to move the sliders under the model chart and view the effect on prediction accuracy. (Hint: it is really easy to achieve 100% accuracy, the purpose is to explore how changing the different model parameters affect training and validation.)

The model uses XGBoost algorithm to predict if a mushroom is edible or poisonous. The baseline is based on the most frequent feature in the training set. Top 10 features are listed to the right of the chart. The data is sourced from the UCI Machine Learning repository. (If your chart appears compressed, try resizing the browser window to knock it back into shape!)