mnist1d_clf_1v7

A binary classifier that tells apart the digits 1 and 7 in the MNIST-1D dataset.

Model

A scikit-learn pipeline: StandardScaler followed by LogisticRegression(max_iter=1000).

Data

Trained on prteek/mnist1d_f: the y == 1 and y == 7 rows of the christopher/mnist1d train split (800 rows). Each example has 40 features (x_0 … x_39) and a label y.

80% of the rows were used for training and 20% were held out for testing, split stratified on y with random_state=42.

Results

Metric	Held-out test set (160 rows)
Accuracy	0.887

Usage

import joblib
from huggingface_hub import hf_hub_download

model = joblib.load(hf_hub_download("prteek/mnist1d_clf", "model.joblib"))
preds = model.predict(X)  # X: DataFrame with columns x_0 … x_39; returns 1 or 7

Limitations

• Only distinguishes 1 from 7; other digits get labelled as one of the two.
• Accuracy comes from a single 160-row test split, so it is a rough estimate.
• joblib files can run arbitrary code when loaded, so only load models you trust.