Nice to Meet You

How I got started

Machine learning is easy

Six Simple Steps

  1. Get data

  2. Prep data

  3. Pick model

  4. Fit model

  5. Validate model

  6. Deploy

GridSearchCV

Pipelines

Informed machine learning is hard

Anscombe's Quartet

Machine Learning in color?

Follow the Yellow Brick Road

Feature Analysis

Model Selection

Hyperparameter Tuning

Boxplots

Histograms

Sploms

Jointplots

Radviz

Parallel Coordinates

Choosing the Right Estimator

Cluster Comparison

Classifier Comparison

Models: Families, Forms, Fitted

Classification Heatmap

ROC-AUC Plots

Prediction Error Plots

Residual Plots

Blind Gridsearch

Validation Curves

Visual Gridsearch

Yellowbrick

Visual Transformers

Search for Separability

Explore Pairs of Features

Evaluate Your Sklearn Models

Identify Potential Problems

Testers Wanted!

bit.ly/ybtesters