Objective:

The objective of this special session is to bring together researchers who apply computational intelligence techniques in time series forecasting. The aim is to exchange their ideas and approaches, to discuss and to present latest results on this emerging field. Statistical techniques are well established in forecasting while, generally much younger, computational techniques provide us promissing comparable results concerning the forecasting accuracy and they also may provide us with some additional value, e.g., robustness, interpretability, transparency, flexibility, low computational efforts or an ability to identify the adequat input variables (lags).

Topics (suggested but not limited to):

• CI (fuzzy, NN, SVM, evolutionary, hybrid etc.) models for time series predictions
• time series feature extraction using CI models,
• robustness of forecasting models,
• linguistic aspects of time series analysis,
• time series trend extraction and the related trend models,
• time series classification and segmentation,
• forecasting of multi-variate time series using CI models,
• forecasting of events and quantitites not related to time series.

Relation to a competition:

We follow the idea of promoting the computatinoal intelligence methods for forecasting based on their development and further comparison with each other as well as with statistical benchmarks. The ideas has been sufficiently propagated by the series of Artificial Neural Network and Computational Intelligence Forecasting competitions such as NN3, NN5, or NNGC1. The CIF special session is connected to the CIF International Competition in Time Series Forecasting. The participation in the competition is not a condition for a submission of papers in the special session. However, we strongly encourage the authors to consider the participation in the competition too.

We look forward to your submissions and also to meet you in Gijon.

Kind regards & good luck!
Martin Štěpnička, Paulo Cortez and Juan Peralta Donate.

We proudly acknowledge the endorsement of CIF under the wings of EUSFLAT Working Group on Learning and Data Mining.