Return to search

Forecasting Global Temperature Variations by Neural Networks

Global temperature variations between 1861 and 1984 are forecast usingsregularization networks, multilayer perceptrons and linearsautoregression. The regularization network, optimized by stochasticsgradient descent associated with colored noise, gives the bestsforecasts. For all the models, prediction errors noticeably increasesafter 1965. These results are consistent with the hypothesis that thesclimate dynamics is characterized by low-dimensional chaos and thatsthe it may have changed at some point after 1965, which is alsosconsistent with the recent idea of climate change.s

Identiferoai:union.ndltd.org:MIT/oai:dspace.mit.edu:1721.1/7208
Date01 August 1994
CreatorsMiyano, Takaya, Girosi, Federico
Source SetsM.I.T. Theses and Dissertation
Languageen_US
Detected LanguageEnglish
Format11 p., 342101 bytes, 403018 bytes, application/octet-stream, application/pdf
RelationAIM-1447, CBCL-101

Page generated in 0.0022 seconds