MindsDB, an open-source applied machine learning platform, announced full integration with Nixtla, an open-source ecosystem that offers time-series forecasting.
Time-series forecasting refers to making scientific predictions based on historical, time-stamped data. It allows data scientists to employ models to predict a future value or classification at a particular point in time, such as forecasting power demand, call volumes, inventory requirements, or supply and demand.
Nixtla offers libraries specifically for time-series forecasting. One of the libraries, StatsForecast, which provides statistical and econometric models, will now function seamlessly within the MindsDB ecosystem. This integration will allow developers using MindsDB to build AI-powered forecasting capabilities and anomaly detection solutions in the database without writing extensive code. MindsDB turns a team of 1,000 developers into 1,000 AI developers with little to no training.
The Nixtla integration includes accurate model implementations, probabilistic forecasting and confidence intervals, support for exogenous variables and static covariates, anomaly detection and time series forecasting. Nixtla’s StatsForecast is optimized for high performance and scalability and uses classical methods, such as ARIMA, rather than deep learning models. This platform means models can be trained quickly and generalized well, making short-time series forecasting easier for developers.