Not supported for mobile device.
Please use the website on a desktop or larger screen.
Tune hyperparameters, watch your model train in real-time, and earn stars based on accuracy. No coding required, just pure intuition. Current phase just simulates training in the browser, Python backend for real model training will update in the future!
⚡
Live Training Simulation
🎛️
Interactive Hyperparameters
⭐
Star Rating System
📊
Real Dataset Challenges
6 challenges across 3 difficulty levels
Beginner
2 challenges
🚢
BEGINNERTitanic Survival
Predict whether a passenger survived the Titanic disaster based on features like age, sex, and passenger class.
FEATURES
23
SAMPLES
891
TARGET
80%
🌸
BEGINNERIris Species
Classify iris flowers into 3 species based on sepal and petal measurements. A classic multi-class classification task.
FEATURES
5
SAMPLES
150
TARGET
82%
Intermediate
2 challenges
🔢
INTERMEDIATEMNIST Digits
Recognize handwritten digits (0-9) from 28×28 pixel grayscale images. The benchmark dataset for image classification.
FEATURES
20
SAMPLES
60K
TARGET
85%
🏠
INTERMEDIATEBoston Housing
Predict median house values using socioeconomic features. Tests your regression tuning skills to minimize error.
FEATURES
19
SAMPLES
506
TARGET
82%
Expert
2 challenges
💳
EXPERTCredit Card Fraud
Detect fraudulent transactions in a highly imbalanced dataset. Requires careful hyperparameter tuning for precision.
FEATURES
20
SAMPLES
285K
TARGET
83%
🎭
EXPERTSentiment Analysis
Classify movie reviews as positive or negative. NLP challenge requiring the right algorithm and regularization strategy.
FEATURES
19
SAMPLES
25K
TARGET
83%
All training is simulated in the browser.