Lab 7: Evaluating Models
Due Wednesday, November 4th by 11:59pm

Learning Goals

• Learn to read and use the scikit-learn API Reference and User's Guide.
• Understand how to load and use scikit-learn python libraries to train and evaluate MLPClassifier models.
• Understand how to reproduce experiments with a specific random seed.
• Explore how to the tune hyperparameters of a neural network, such as the hidden layer configuration.
• Analyze experiments using several methodologies, including basic statistics, confusion matrices and validation curves.

Starting point code

This lab will be done with a partner. You and your partner should have a github repo ready to go by the time lab starts. Please let the instructor know if you have any issues accessing it.

Introduction

In a previous lab you implemented neural networks from scratch. In this lab you learn how to build neural network models using the python library called scikit-learn. We will focus on exploring neural network models in more depth, applying them to two different datasets. In scikit-learn the basic neural network model is called MLPClassifier where the MLP stands for Multi-Layer Perceptron.

All the instructions for the lab are included in the Jupyter Notebook eval-models.ipynb.

Submitting

When you are finished, use git to add, commit, and push your work.