Chronos : A pre-trained time series forecasting model based on a language model architecture

Chronos

AI Model AI Time Series Forecasting #Time Series Forecasting #Probabilistic Prediction #Pre-trained Model #Machine Learning Standard Picks Open Source

Overview :

Chronos is a series of pre-trained time series forecasting models based on language model architectures. Time series are converted into sequences of tokens through scaling and quantization, and then a language model is trained using cross-entropy loss. After training, multiple future trajectories are sampled given the historical context to obtain probabilistic predictions. The Chronos model has been trained on a large amount of publicly available time series data and synthetic data generated using Gaussian processes.

Target Users :

Suitable for researchers and developers who need to conduct time series forecasting, such as in fields like financial market analysis, sales forecasting, and resource planning.

Total Visits： 474.6M

Top Region： US(19.34%)

Website Views ： 76.7K

Use Cases

Financial analysts use the Chronos model to predict stock price trends

Retailers use Chronos for sales forecasting to optimize inventory management

Energy companies use the Chronos model to predict electricity demand, enabling more efficient resource allocation

Features

Converts time series into token sequences

Trains a language model using cross-entropy loss

Obtains probabilistic predictions through sampling