Anticipated-Learning-Machine

Dataset

Coupled Time-Variant Lorentz System

This is a synthetic 90-dimension time-variant coupled Lorentz system.

Gene Dataset of Rats

This 31099-dimension dataset is based on gene expression profiles with Affimetrix microarray measured on the laboratory rat (Rattus norvegicus) cultured cells from SCN.

Plankton Dataset

This 58-dimension dataset is collected from an optical plankton counter (OPC) and CTD mounted to a ScanFish platform that was towed and undulated behind the vessel.

Ground Ozone Level Dataset

This 72-dimension dataset is collected from 1998 to 2004 at the Houston, Galveston and Brazoria area. Each variable represents a property of the observed area.

Wind Dataset

This 155-dimension dataset is collected in Japan by the Japan Methodological Agency. Each variable represents the wind speed at different wind station.

Stock Index Dataset

This 1130-dimension dataset is collected from 2018-05-01 to 2018-11-22 of different 1130 stock indexes with an interval of 1 day except Saturday and Sunday. Each stock index is a relative number of stock price statistics that measure and reflect the overall price level of the stock market and its changing trend.

Traffic Dataset

This 207-dimension dataset contains traffic speed collected from 207 loop detectors in the highway of Los Angeles County. Each variable represents the traffic flow at different detectors.

Satellite cloud image dataset

This 241-dimensional dataset comes from the satellite cloud image data recording the route of typhoon Marcus collected by National Institute of Informatics. Each variable represents the cloud image.

Experiment Result

Requirements

• python 3.7
• pytorch 1.2.0 (important)
• numpy 1.16.4
• CUDA Version 10.2 (important)
• GPU & linux

Note that the version of pytorch and CUDA are important factors and the results vary slightly depending on the device.

It is noteworthy that the hyper-parameters setting remains to be sensitive in some of the experiments under the current framework. This is mainly due to strong nonlinearity or/and stochasticity of the dynamical systems also with the observed noisy data, and thus how to make more in-depth theoretical analysis and further develop an appropriate framework taking these issues into consideration is an open and interesting problem in future.