- author = "Stephen Kelly and Jacob Newsted and Wolfgang Banzhaf and Cedric Gondro",
- title = "A Modular Memory Framework for Time Series Prediction",
- year = "2020",
- editor = "Carlos Artemio {Coello Coello} and Arturo Hernandez Aguirre and Josu Ceberio Uribe and Mario Garza Fabre and Gregorio {Toscano Pulido} and Katya Rodriguez-Vazquez and Elizabeth Wanner and Nadarajen Veerapen and Efren Mezura Montes and Richard Allmendinger and Hugo Terashima Marin and Markus Wagner and Thomas Bartz-Beielstein and Bogdan Filipic and Heike Trautmann and Ke Tang and John Koza and Erik Goodman and William B. Langdon and Miguel Nicolau and Christine Zarges and Vanessa Volz and Tea Tusar and Boris Naujoks and Peter A. N. Bosman and Darrell Whitley and Christine Solnon and Marde Helbig and Stephane Doncieux and Dennis G. Wilson and Francisco {Fernandez de Vega} and Luis Paquete and Francisco Chicano and Bing Xue and Jaume Bacardit and Sanaz Mostaghim and Jonathan Fieldsend and Oliver Schuetze and Dirk Arnold and Gabriela Ochoa and Carlos Segura and Carlos Cotta and Michael Emmerich and Mengjie Zhang and Robin Purshouse and Tapabrata Ray and Justyna Petke and Fuyuki Ishikawa and Johannes Lengler and Frank Neumann",
- isbn13 = "9781450371285",
- publisher = "Association for Computing Machinery",
- publisher_address = "New York, NY, USA",
-
URL = "
https://doi.org/10.1145/3377930.3390216",
-
DOI = "
doi:10.1145/3377930.3390216",
- booktitle = "Proceedings of the 2020 Genetic and Evolutionary Computation Conference",
- pages = "949--957",
- size = "9 pages",
- keywords = "genetic algorithms, genetic programming, Linear genetic programming, LGP, memory, time series prediction, modularity",
- address = "internet",
- series = "GECCO '20",
- month = jul # " 8-12",
- organisation = "SIGEVO",
- abstract = "Tangled Program Graphs (TPG) is a framework for genetic programming which has shown promise in challenging reinforcement learning problems with discrete action spaces. The approach has recently been extended to incorporate temporal memory mechanisms that enable operation in environments with partial-observability at multiple timescales. Here we propose a highly-modular memory structure that manages temporal properties of a task and enables operation in problems with continuous action spaces. This significantly broadens the scope of real-world applications for TPGs, from continuous-action reinforcement learning to time series forecasting. We begin by testing the new algorithm on a suite of symbolic regression benchmarks. Next, we evaluate the method in 3 challenging time series forecasting problems. Results generally match the quality of state-of-the-art solutions in both domains. In the case of time series prediction, we show that temporal memory eliminates the need to pre-specify a fixed-size sliding window of previous values, or autoregressive state, which is used by all compared methods. This is significant because it implies that no prior model for a time series is necessary, and the forecaster may adapt more easily if the properties of a series change significantly over time.",
-
notes = "Nominated for Best Paper.
Teams of programs, read only regsiters, R/W memory, population. Tangled program graphs (TPG). Laser, sunspot, Mackey-Glass source code available
Also known as \cite{10.1145/3377930.3390216} GECCO-2020 A Recombination of the 29th International Conference on Genetic Algorithms (ICGA) and the 25th Annual Genetic Programming Conference (GP)",
