Dark Mode
Pendigits Classification Data
Data Science and Analytics
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
Explore and train advanced models for handwritten digit classification using this robust collection of pen-based samples. The data provides a solid foundation for developing and evaluating high-performance handwriting recognition systems, making it highly valuable for machine learning projects focused on classification tasks.
Columns
The dataset contains 17 columns in total.
- input1 - input16: These 16 columns represent the numerical features capturing the normalized (x, y) coordinates of the pen-tip movement during the writing of a single digit.
- class: The target column, defining the digit label (0 through 9) that the pen movement represents.
Distribution
This dataset is structured with 10,992 total samples, each representing a single handwritten digit observation. The data is supplied as a CSV file (
pendigits_txt.csv), featuring 17 columns per record. Each row is fully valid, with zero mismatched or missing values, ensuring immediate usability for modelling projects.Usage
The data is ideally suited for several key applications:
- Training and evaluating machine learning algorithms designed for handwritten digit classification.
- Benchmarking performance across various models, including Support Vector Machines (SVM), Decision Trees, and Neural Networks.
- Conducting exploratory data analysis and visualising handwriting patterns to inform feature engineering techniques.
Coverage
This dataset focuses strictly on technical pen movement features and digit labels. It does not include geographical, temporal, or specific demographic metadata, as it is designed purely for pattern recognition tasks related to classification.
License
CC BY-NC-SA 4.0
Who Can Use It
Intended users include:
- Data Scientists and Machine Learning Engineers: Utilising the dataset to develop, test, and benchmark models for optical character recognition (OCR) and handwriting recognition systems.
- Academic Researchers: Exploring advanced pattern recognition techniques and novel feature engineering strategies.
- Students and Educators: Learning about core classification challenges and experimenting with fundamental machine learning algorithms.
Dataset Name Suggestions
- Pendigits Classification Data
- Handwritten Digit Trajectory Samples
- Pen-Based Digit Movement Features
- Handwriting Recognition Benchmark Dataset
Attributes
Original Data Source: Pendigits Classification Data
QUALITYLoading...