ask - Online in the Cloud

PROGRAM:

NAME

ask - Adaptive Sampling Kit, a toolkit for sampling big experimental spaces

DESCRIPTION

ASK stands for Adaptive Sampling Kit. Its main goal is accelerating long experiments. Two
kinds of variables define an experiment: a response and a set of factors. The response is
characterized on the factors' Cartesian product.

When the space is small, the response can be measured for every point in the space. When
the space is large, doing an exhaustive measurement is either not possible in terms of
execution time or simply not practical. ASK tries to find good approximations of the
response by sampling only a small fraction of the space. ASK features multiple active
learning algorithms to prioritize the exploration of the interesting parts of the
experimental space.

SYNOPSIS

ask [-h] [--force_overwrite] [--replay_only] [--skip_model]
[--skip_reporter] [--version] configuration

COMMAND LINE OPTIONS

configuration
the experiment configuration file

-h, --help
show this help message and exit

--force_overwrite
If a previous output directory is found its contents will be overwritten

--replay_only
(Must be run over a full previous ask run). Skip bootstrap and sampler modules, but
runs previous labels through the model and reporter. This option is useful if one
wants to run sampling and analyze results separately.

--skip_model
Skip model module. (WARNING: if the chosen control module relies on the output of
the model module, this option fails.)

--skip_reporter
Skip reporter module. (WARNING: if the chosen control module relies on the output
of the reporter module, this option fails.)

--version
show program's version number and exit

COPYRIGHT

Copyright (C) 2011-2012 Universite de Versailles St-Quentin-en-Yvelines

The Adaptive Sampling Kit (ASK) is released under the GNU General Public License, version
2. Please refer to the file LICENSE that accompanies your copy of ASK.

Use ask online using onworks.net services