The NN3 competition was designed to evaluate how modern neural network (NN) and computational intelligence (CI) methods compare to traditional statistical benchmarks like those used in the M3 competition. Composition:
Includes a wide range of real-world business patterns (e.g., industry data), making it a robust test for model generalization.
For modern use, researchers often access these files through the NN3 Official Website or data science repositories like Kaggle . Critical Reception nn3.zip
The historical data is typically provided in vertical columns of varying lengths.
The "masked" nature of the data (anonymized origin) ensures that models must rely on time series patterns rather than domain-specific knowledge. Practical Considerations The NN3 competition was designed to evaluate how
111 monthly time series, including the 11 from the reduced set.
It is a standard historical benchmark in the forecasting community and is often included in modern research packages like the tscompdata R package on GitHub. Critical Reception The historical data is typically provided
The series vary in length (68 to 144 observations) and include seasonal, non-seasonal, and "difficult" patterns with outliers and structural breaks. Key Strengths