TECHNIQUES FOR COMPRESSING A LARGE DISTRIBUTED EMPIRICAL SAMPLE OF A COMPOUND PROBABILITY DISTRIBUTION INTO AN APPROXIMATE PARAMETRIC DISTRIBUTION WITH SCALABLE PARALLEL PROCESSING

摘要

Techniques for estimated compound probability distribution are described. An apparatus may comprise a configuration component, perturbation component, sample generation controller, an aggregation component, a distribution fitting component, and statistics generation component. The configuration component may be operative to receive a compound model specification and candidate distribution definition. The perturbation component may be operative to generate a plurality of models from the compound model specification. The sample generation controller may be operative to initiate the generation of a plurality of compound model samples from each of the plurality of models. The distribution fitting component may generate parameter values for the candidate distribution definition based on the compound model samples. The statistics generation component may generate approximated aggregate statistics. Other embodiments are described and claimed.

主权项

1. At least one computer-readable storage medium comprising instructions that, when executed, cause a system to:
receive a plurality of model samples, the model samples implying a non-parametric distribution of aggregate loss events; receive a candidate distribution definition, the candidate distribution definition comprising a combination of at least two component distributions, the candidate distribution definition comprising one or more parameters; determine parameter values for the one or more parameters of the candidate distribution definition, the parameter values determined by optimizing a non-linear objective function through a search over a multidimensional space of parameter values, the optimization performed by a distribution fitting component operating on a processor circuit, the objective function calculating a distance between the non-parametric distribution of the aggregate loss events as implied by the model samples and a parametric distribution determined by application of potential parameter values to the candidate distribution definition; generate approximated aggregate statistics for the plurality of model samples based on an optimized parametric distribution defined by the candidate distribution definition and the determined parameter values; and report the approximated aggregate statistics.