A/B Test Simulator - Conversion Rate Prediction
Train a logistic regression model on A/B experiment observations and use what-if sliders to predict conversion probability.
How to use
- 1
Paste experiment JSON
Upload a CSV (header row + 0/1 label column) or paste the JSON format directly. Click 'Load example' to see the expected structure.
- 2
Train the model
Logistic regression is fitted via gradient descent in the browser.
- 3
Use what-if sliders
Adjust feature values to predict conversion probability for hypothetical scenarios.
Why use this tool?
- Fits logistic regression entirely client-side via WebAssembly
- What-if sliders enable instant scenario exploration
- No data leaves your browser - safe for sensitive experiment data
Frequently asked questions
- What input format is expected?
- A CSV file where columns are features and the last column (or any column named 'label', 'target', or 'converted') is the 0/1 outcome. The tool also accepts raw JSON: {features: [[...]], labels: [...], feature_names: [...]}.
- How many iterations does training run?
- 500 gradient descent iterations with a learning rate of 0.1. Sufficient for most linearly-separable experiments.
Related tools
Provide match and non-match string examples; a decision tree learns the pattern and exports it as Rust, JavaScript, or regex.
Upload a CSV to auto-detect column types, spot missing values, and fit a linear regression - all locally in your browser.
Popular right now
Format, validate, and minify JSON instantly in your browser. Your data never leaves your device.
Decode JWT tokens and inspect header and payload instantly in your browser. Your tokens never leave your device.
Count words, characters, sentences, and estimate reading time instantly in your browser. No sign-up required.