CCBias: An Event Detection Optimization Framework

Loading...
Thumbnail Image

Date

2019

Authors

Faridani, Theo H.

Journal Title

Journal ISSN

Volume Title

Publisher

University of Oregon

Abstract

We present a software, CCBias, to assist researchers in observing events of all kinds. Given characteristic information about a population of objects and observational methods, CCBias can generate synthetic observational data. CCBias can also recommend search strategies if told what observational outcomes are desirable. Lastly, CCBias can estimate the bias in real data by transforming the problem of identifying bias into a problem of estimating model parameters. We demonstrate the strengths and weaknesses of CCBias in a case study focused on planetary defense. CCBias is written in the Python programming language.

Description

35 pages

Keywords

Physics, Observation, Computer Science, Censored Data, Bias, Data Science

Citation