# MathFi.ai

## Docs

- [Login](https://docs.mathfi.ai/api-reference/authentication/login.md): Authenticates a user and returns a JWT Bearer token. The token is valid for **1 hour** and must be included as `Authorization: Bearer <token>` on all `/api/v1/*` requests.
- [Create dataset](https://docs.mathfi.ai/api-reference/datasets/create.md): Creates a new empty dataset. Dataset creation is a two-step process:
- [Delete dataset](https://docs.mathfi.ai/api-reference/datasets/delete.md): Permanently deletes a dataset and its associated data.
- [Get dataset](https://docs.mathfi.ai/api-reference/datasets/get.md): Retrieve dataset details by key. Poll this endpoint to check when status reaches `COMPLETED`.
- [List datasets](https://docs.mathfi.ai/api-reference/datasets/list.md): Returns a paginated list of all datasets for the authenticated user.
- [Get model](https://docs.mathfi.ai/api-reference/models/get.md): Retrieve details and performance metrics for a trained model.
- [List models by dataset](https://docs.mathfi.ai/api-reference/models/list-by-dataset.md): Returns all models trained against a given dataset, paginated. Models are versioned — each successful training run that improves on the previous creates a new version.
- [Create prediction](https://docs.mathfi.ai/api-reference/predictions/create.md): Submits an unseen (unlabelled) CSV file for inference using the specified model.
- [Get prediction](https://docs.mathfi.ai/api-reference/predictions/get.md): Retrieve prediction details including the download URL for the result CSV.
- [List predictions by model](https://docs.mathfi.ai/api-reference/predictions/list-by-model.md): Returns all predictions run using a given model, paginated.
- [Get prediction progress](https://docs.mathfi.ai/api-reference/predictions/progress.md): Poll this endpoint to check prediction status. When status is `COMPLETED`, retrieve the result via `GET /api/v1/predictions/{predictionKey}`.
- [Cancel training job](https://docs.mathfi.ai/api-reference/training/cancel.md): Cancels a training job that is in `PENDING` or `RUNNING` status.
- [Start training](https://docs.mathfi.ai/api-reference/training/create.md): Creates a new training job for a dataset in `COMPLETED` status.
- [Get training job](https://docs.mathfi.ai/api-reference/training/get.md): Retrieve details and current status of a training job.
- [List training jobs](https://docs.mathfi.ai/api-reference/training/list-by-dataset.md): Returns all training jobs for a given dataset, paginated.
- [Get training progress](https://docs.mathfi.ai/api-reference/training/progress.md): Returns the real-time progress of all competing algorithms in a training job.
- [API Reference Overview](https://docs.mathfi.ai/developers/api-overview.md): MathFi.ai REST API reference covering Bearer token auth, base URL, and endpoint structure for datasets, training, and inference.
- [API Recipes](https://docs.mathfi.ai/developers/api-recipes.md): API workflow examples for MathFi.ai covering authentication, dataset creation, model training, predictions, and hyperparameter tuning with curl.
- [Authentication](https://docs.mathfi.ai/developers/authentication.md): Authenticate with the MathFi.ai API. Obtain a Bearer token, handle 1-hour expiry, and secure your requests.
- [Python SDK](https://docs.mathfi.ai/developers/python-sdk.md): MathFi.ai Python SDK for installing from source, authenticating, and listing datasets. Early access, full lifecycle via REST API.
- [Glossary](https://docs.mathfi.ai/glossary.md): Key terms and metrics for MathFi.ai, including precision, recall, F1, AUC, overfitting, and binary vs multi-class classification.
- [Hyperparameter tuning](https://docs.mathfi.ai/guides/hyperparameter-tuning.md): Tune MathFi.ai models for better accuracy by adjusting numberOfBuckets, performanceThreshold, and scalingFactor.
- [Input CSV File Format](https://docs.mathfi.ai/guides/input-csv-creation.md): How to format your labelled CSV data for MathFi.ai. Covers column requirements, target variable setup, and data preparation tips.
- [Training](https://docs.mathfi.ai/guides/training-guide.md): Step-by-step guide to training predictive models on MathFi.ai, covering dataset creation, model training, and running predictions.
- [FAQ](https://docs.mathfi.ai/help/faq.md): Frequently asked questions about MathFi.ai covering implementation time, supported problem types, data security, pricing, and API integration.
- [MathFi.ai Documentation](https://docs.mathfi.ai/index.md): MathFi.ai docs: build production predictive AI models from CSV data in minutes on commodity CPUs. REST API, Python SDK, guides, and use cases.
- [Getting started](https://docs.mathfi.ai/quickstart.md): Get started with MathFi.ai. Create your first dataset, train a model, and run predictions in under an hour.
- [Agriculture crop classification](https://docs.mathfi.ai/use-cases/agriculture.md): Agriculture classification use case to predict crop or soil outcomes from structured field data with MathFi.ai.
- [Banknote forgery detection: binary classification](https://docs.mathfi.ai/use-cases/banknote-forgery.md): Banknote forgery detection: binary classification of authentic vs counterfeit notes using MathFi.ai.
- [Credit card approval prediction: 95% accuracy](https://docs.mathfi.ai/use-cases/credit-card-approval.md): Credit card approval prediction achieving 95% accuracy on application decisions using MathFi.ai.
- [Healthcare outcome prediction](https://docs.mathfi.ai/use-cases/healthcare.md): Healthcare prediction use case: classify patient outcomes using MathFi.ai balanced-class learning on tabular clinical data.
- [Insurance claims prediction: imbalanced data](https://docs.mathfi.ai/use-cases/insurance-claim.md): Insurance claims prediction to classify claim outcomes on imbalanced data with MathFi.ai balanced-class learning.
- [Self-checkout fraud detection](https://docs.mathfi.ai/use-cases/self-checkout-fraud.md): Self-checkout fraud detection to identify fraudulent retail transactions with balanced-class ML on tabular data.

## OpenAPI Specs

- [openapi](https://docs.mathfi.ai/api-reference/openapi.yaml)