ArinRoths/shoe-data

Shoe Dataset

ArinRoths/shoe-data

This dataset contains measurements and categorical features collected from 30 shoes. The prediction target is the roommate who owns each shoe.

Source and task

The original data was manually collected from shoes and includes five predictor features: shoe length, shoe height, type, color, and closure. The target is Roomate.

The 30 original samples were split into training, validation, and test sets before augmentation. Only original training samples were used to generate synthetic data.

Fields

Field	Meaning
Length(in)	Shoe length in inches
Height(in)	Shoe height in inches
Type	Type of shoe
Color	Shoe color
Closure	Shoe closure type
Roomate	Prediction target: owner of the shoe
source_id	Unique row identifier
parent_id	Original training source for the row
augmentation	Augmentation method used
is_augmented	Whether the row is synthetic

Splits

Split	Original rows	Synthetic rows	Total rows
train	21	343	364
validation	4	0	4
test	5	0	5

Original samples were randomly split before augmentation using seed 24679. Validation and test samples remain unchanged.

Augmentation and lineage

This run requests 6 copies per method and original training row before unchanged and duplicate samples are removed.

• Additive numeric jitter: adds small random variation to shoe length and height.
• Multiplicative numeric scaling: scales shoe length and height by a small random percentage.
• Categorical perturbation: changes one of type, color, or closure to another value observed in the training data.

The roommate target remains unchanged during augmentation. Synthetic samples retain their original training source in parent_id, while augmentation records the method used and is_augmented distinguishes synthetic rows from originals.

Training method	Stored rows
additive_numeric_jitter	125
categorical_perturbation	92
multiplicative_numeric_scale	126
none	21

Validation and test rows remain original and are never used to generate synthetic samples.

Intended use and limitations

This dataset was created for a classroom exercise in tabular data preparation and augmentation. It can be used to explore a classification task in which shoe measurements and categorical features are used to predict the roommate who owns the shoe.

The dataset is small and contains shoes from a limited number of people, so it is not representative of shoes or shoe owners in general. Synthetic samples are variations of the original training data and do not represent additional independently collected shoes.

Privacy and licensing

The dataset contains measurements and categorical descriptions of shoes and the roommate who owns each shoe. It does not contain faces, contact information, or other direct personal identifiers. Source IDs are used to track original and synthetic samples and do not contain identifying information. The raw CSV remains a separate source artifact.

No license was specified in repository metadata when this card was first added; this card does not assign one.

Load and compare

from datasets import load_dataset
ds = load_dataset("ArinRoths/shoe-data")
# Train with ds["train"], choose settings with ds["validation"], then score ds["test"].

Use an account with access if repository visibility changes. For reproducible comparisons, record the dataset commit and model/environment versions. Regenerate this card with the preparation notebook after changing the data; its counts are calculated from the actual packaged splits. The YAML schema and split configuration are preserved from the upload.

AI Usage Disclosure

I used ChatGPT to help understand the assignment reqirements and how the code from last week would need to changed based on the data set I collected. This included learned some new functions and replacing others slightly. I also used it to debug my code.