{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/forecasting-across-time-series-databases","title":"Forecasting Across Time Series Databases using Recurrent Neural Networks on Groups of Similar Series: A Clustering Approach","arxiv_id":"1710.03222","date":"2017-10-09","proceeding":null,"authors":["Kasun Bandara","Christoph Bergmeir","Slawek Smyl"],"abstract":"With the advent of Big Data, nowadays in many applications databases\ncontaining large quantities of similar time series are available. Forecasting\ntime series in these domains with traditional univariate forecasting procedures\nleaves great potentials for producing accurate forecasts untapped. Recurrent\nneural networks (RNNs), and in particular Long Short-Term Memory (LSTM)\nnetworks, have proven recently that they are able to outperform\nstate-of-the-art univariate time series forecasting methods in this context\nwhen trained across all available time series. However, if the time series\ndatabase is heterogeneous, accuracy may degenerate, so that on the way towards\nfully automatic forecasting methods in this space, a notion of similarity\nbetween the time series needs to be built into the methods. To this end, we\npresent a prediction model that can be used with different types of RNN models\non subgroups of similar time series, which are identified by time series\nclustering techniques. We assess our proposed methodology using LSTM networks,\na widely popular RNN variant. Our method achieves competitive results on\nbenchmarking datasets under competition evaluation procedures. In particular,\nin terms of mean sMAPE accuracy, it consistently outperforms the baseline LSTM\nmodel and outperforms all other methods on the CIF2016 forecasting competition\ndataset.","url_abs":"http://arxiv.org/abs/1710.03222v2","url_pdf":"http://arxiv.org/pdf/1710.03222v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"forecasting-across-time-series-databases","repo_url":"https://github.com/EvgeniyaMartynova/MLiP_M5","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"forecasting-across-time-series-databases","repo_url":"https://github.com/arsalan993/Multivariate-timeseries-forecasting-using-TS-decompostion-and-deep-learnring","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null},{"paper_slug":"forecasting-across-time-series-databases","repo_url":"https://github.com/arsalan993/Univariate-Time-Series-Forecasting-using-decomposition-and-Deep-Learning","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"benchmarking","task_name":"Benchmarking"},{"task_slug":"clustering","task_name":"Clustering"},{"task_slug":"time-series-1","task_name":"Time Series"},{"task_slug":"time-series","task_name":"Time Series Analysis"},{"task_slug":"time-series-clustering","task_name":"Time Series Clustering"},{"task_slug":"time-series-forecasting","task_name":"Time Series Forecasting"},{"task_slug":"univariate-time-series-forecasting","task_name":"Univariate Time Series Forecasting"}],"methods":[{"method_slug":"lstm","method_name":"LSTM"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}