ELA Feature Values & Statistics

Values and statistics for 33 ELA features:- FDC- Dispersion [7 levels log10 spaced]- Information significance [First, Second and N orders]- Information Content [Using both NN and Random sorting]- Linear and Quadratic meta-models- Skewness, Kurtosis and Entropy of cost distributionStatistics were calculated through bootstrapping the function data (composed of Xinput and Yinput) using bootstrap indexes. These are:- Bootstrap variance- Mean Squared Error [defined as the squared difference between the value and the bootstrap mean]- Bias- Confidence Interval at 95% [Divided into High and Low value]Raw bootstrap values are provided in a different dataset. Code to be uploaded.If you use this dataset please use the citation for the complete collection, as follows:Mario A. Muñoz "LEOPARD: LEarning and OPtimization Archive of Research Data, Version 1.0" Aug. 2020 DOI:10.6084/m9.figshare.c.5106758.