Sparkle is a Programming by Optimisation (PbO)-based problem-solving platform designed to enable the widespread and effective use of PbO techniques for improving the state-of-the-art in solving a broad range of prominent AI problems, including SAT and AI Planning.
Features describing instances are usually split into feature groups. It is then possible to decide which feature groups to extract and save time by extracting only the necessary feature groups.
AutoFolio supports feature groups, however only with ASlib scenario as an input.
Currently, Sparkle constructs the portfolio selector with CSV files.
To improve the performance we need to:
Allow the feature extractor to specify feature groups
Extract the features based on separate feature groups, and measure the extraction time accordingly
Pass the feature groups to the selector constructor
Features describing instances are usually split into feature groups. It is then possible to decide which feature groups to extract and save time by extracting only the necessary feature groups. AutoFolio supports feature groups, however only with ASlib scenario as an input. Currently, Sparkle constructs the portfolio selector with CSV files. To improve the performance we need to: