01 Introduction

Material

Chapter 1 + the pdf "kNN"

kNN (PDF)

Lesson 1 (PDF)

Titanic (CSV)

Titanic (Jupyter Notebook)

Session Description

We will talk about what machine learning is and the types of problems we work with. We will also introduce the first algorithm: k-Nearest Neighbor.

Key Concepts

• Distance metrics in kNN
• The parameter “k” and its effect on model performance
• Differences between low bias and low variance models
• Data normalization for distance-based methods
• Evaluating model outputs (accuracy, confusion matrix, etc.)

Learning Objectives

• Explain what is meant by the term Machine Learning (ML)
• Explain what is meant by supervised vs. unsupervised learning
• Explain the overall difference between classification and regression
• Describe the "train-test" methodology
• Train a k-Nearest Neighbors (kNN) algorithm on a dataset in sklearn
• Explain the principles behind the kNN algorithm
• Explain what is meant by the "hyperparameters" of an algorithm