Machine Learning

Project 4

This page is currently being updated to reflect the instructions for Fall 2025. There are items that may not be fully updated yet

Project Description

In this project, you will use various machine learning algorithms for image classification on the MNIST dataset. Then, you will use the best algorithm to perform image classification on the robot in a museum tour guide task.

  • Part 1: Nearest Neighbors in Python
  • Part 2: Training Machine Learning algorithms with SciKit Learn
  • Part 3: Robot Museum Tour Guide

Teams

This project will be completed individually. You will not need a robot until the very end. At that point, we will assign two people per robot.

Getting Your Code

To get your starter code download the files by clicking here and extracting them, or by running the commands below in your terminal.

  • Navigate to a directory that is not already a git repository using cd. If running git status outputs fatal: ... you’re not in a git repository.
  • Download the files from the website wget https://robotics102.org/um-f25/assets/template_code/p4_notebooks_template.zip.
  • Unzip the files unzip p4_notebooks_template.zip.
  • Delete the zip archive rm p4_notebooks_template.zip.

To get the starter code for the robot portion of the project download the files by clicking here and extracting them, or by running the commands below in your terminal.

  • Navigate to a directory that is not already a git repository using cd. If running git status outputs fatal: ... you’re not in a git repository.
  • Download the files from the website wget https://robotics102.org/um-f25/assets/template_code/p4_robot_tour_guide_template.zip.
  • Unzip the files unzip p4_robot_tour_guide_template.zip.
  • Delete the zip archive rm p4_robot_tour_guide_template.zip.

Setting Up Your Repository

You will want to complete the notebooks on Google CoLab by uploading them to your Google Drive.

To set up a repository for the robot code, follow the guide.