Overview

Handwritten digit recognition is a fundamental computer vision task and a classic application of image classification. It plays an important role in applications such as document digitization, postal code recognition, form processing, and optical character recognition (OCR). By learning visual patterns from labeled examples, machine learning models can automatically identify handwritten digits with high accuracy.

In this challenge, your task is to build a machine learning model that recognizes handwritten digits from images. You will work with a labeled image dataset where each image contains a single handwritten digit from 0 to 9. This challenge is beginner-friendly and provides hands-on experience with the complete computer vision workflow, including image preprocessing, data exploration, model training, and evaluation for multi-class image classification.

Practice Skills

In this challenge, you will gain hands-on experience with:

• Python
• Image Preprocessing
• Exploratory Data Analysis (EDA)
• Computer Vision
• Image Classification
• Machine Learning / Deep Learning
• Model Evaluation

Evaluation

Goal

Train an image classification model that predicts the handwritten digit (0–9) for each image in the test set.

Metric

Submissions are evaluated on Accuracy:

$\text{Accuracy} = \frac{\text{Correct Predictions}}{\text{Total Predictions}}$

Submission

Format

A CSV with two columns:

• id: The unique identifier for each image.
• digit: The predicted digit label (0–9).

id	digit
000039d3a3924cba84e8098d6bdd6951.png	0
0019e9a6ca414f7ab1456c0f0b578e64.png	3

See the submission guide for upload instructions.

Tracks

The challenge accepts two submission tracks:

• Public: predict on the public test set and upload the resulting CSV.
• Private: upload your model. It is scored on a held-out private test set. Your code must recursively scan the entire data directory.