General¶
Learn about time series analysis
Chronos is a foundational model for zero-shot probabilistic forecasting of univariate time series. The model converts a time series into a sequence of tokens through scaling and quantization. The scaling procedure divides the time series by its mean absolute value, while the quantization process maps the scaled time series values to a discrete set of tokens using uniform binning…
Forecasting
September 2, 2024
Building a well-performing machine learning model requires substantial time and resources. Automated machine learning (AutoML) automates the end-to-end process of building, training and tuning machine learning models. This not only accelerates the development cycle, but also makes machine learning more accessible to those without specialized data science expertise…
Classification
August 20, 2024
Forecasting commodity prices is a particularly challenging task due to the intricate interplay of supply and demand dynamics, geopolitical factors, and market sentiment fluctuations. Deep learning models have been shown to be more effective than traditional statistical models at capturing the complex and non-linear relationships inherent in commodity markets…
Forecasting
July 26, 2024
Inflation forecasts are used for informing economic decisions at various levels, from households to businesses and policymakers. The application of machine learning methods to inflation forecasting offers several potential advantages, including the ability to handle large and complex datasets, capture nonlinear relationships, and adapt to changing economic conditions…
Forecasting
March 20, 2024
FRED-MD is an open-source dataset of monthly U.S. macroeconomic indicators maintained by the Federal Reserve Bank of St. Louis. The FRED-MD dataset was introduced to provide a common benchmark for comparing model performance and to facilitate the reproducibility of research results…
Datasets
January 11, 2024