Nn3.zip 100%

Contestants were traditionally required to forecast 18 points into the future.

111 monthly time series, including the 11 from the reduced set.

The historical data is typically provided in vertical columns of varying lengths.

The series vary in length (68 to 144 observations) and include seasonal, non-seasonal, and "difficult" patterns with outliers and structural breaks. Key Strengths

A review of typically refers to the dataset from the NN3 Forecasting Competition (2006–2007), a seminal event in neural networks and computational intelligence for time series forecasting. This file usually contains a collection of 111 monthly time series drawn from empirical business data. Dataset Overview

The "masked" nature of the data (anonymized origin) ensures that models must rely on time series patterns rather than domain-specific knowledge. Practical Considerations

For modern use, researchers often access these files through the NN3 Official Website or data science repositories like Kaggle . Critical Reception