Zipcode Analysis
Homework 2 applies classification techniques to real-world handwritten digit images from US postal zipcodes. Compare linear regression and KNN approaches with detailed error analysis.
Learning Objectives
- Apply classification methods to image data
- Compare linear vs. non-linear classifiers
- Analyze misclassification patterns
- Understand feature representation in image classification
ESLII Reference
This homework continues from Chapter 2 (Overview of Supervised Learning), applying concepts to the classic zipcode digit recognition problem discussed in the textbook.
Available Scripts
| Script | Description | Subdirectory |
|---|---|---|
zipcode_analysis.py |
Complete analysis pipeline for digit classification | root |
Quick Start
# CLI exploration
cd domains/Stan/cli
python main.py "homework 2"
# Cockpit GUI
cd domains/Stan/cockpit
python stan_cockpit.py
# Enter: "explore zipcode analysis"
# Direct tool access
from unified_agent import StanDataClient, ToolRegistry
client = StanDataClient()
tools = ToolRegistry(client)
result = tools.get_tool('load_hmk2_info')({})
Related Tools
| Tool | Description |
|---|---|
load_hmk2_info |
Get Homework 2 metadata and available scripts |
list_hmk2_scripts |
List all Python scripts in the Hmk2 directory |
compare_homeworks |
Compare with Homework 1 (both use KNN) |