Distinguishing Dynamical Kinds: An Approach for Automating Scientific Discovery
| dc.contributor.author | Shea-Blymyer, Colin | en |
| dc.contributor.committeechair | Jantzen, Benjamin C. | en |
| dc.contributor.committeemember | Huang, Bert | en |
| dc.contributor.committeemember | Karpatne, Anuj | en |
| dc.contributor.committeemember | Prakash, B. Aditya | en |
| dc.contributor.department | Computer Science | en |
| dc.date.accessioned | 2020-12-24T07:00:21Z | en |
| dc.date.available | 2020-12-24T07:00:21Z | en |
| dc.date.issued | 2019-07-02 | en |
| dc.description.abstract | The automation of scientific discovery has been an active research topic for many years. The promise of a formalized approach to developing and testing scientific hypotheses has attracted researchers from the sciences, machine learning, and philosophy alike. Leveraging the concept of dynamical symmetries a new paradigm is proposed for the collection of scientific knowledge, and algorithms are presented for the development of EUGENE – an automated scientific discovery tool-set. These algorithms have direct applications in model validation, time series analysis, and system identification. Further, the EUGENE tool-set provides a novel metric of dynamical similarity that would allow a system to be clustered into its dynamical regimes. This dynamical distance is sensitive to the presence of chaos, effective order, and nonlinearity. I discuss the history and background of these algorithms, provide examples of their behavior, and present their use for exploring system dynamics. | en |
| dc.description.abstractgeneral | Determining why a system exhibits a particular behavior can be a difficult task. Some turn to causal analysis to show what particular variables lead to what outcomes, but this can be time-consuming, requires precise knowledge of the system’s internals, and often abstracts poorly to salient behaviors. Others attempt to build models from the principles of the system, or try to learn models from observations of the system, but these models can miss important interactions between variables, and often have difficulty recreating high-level behaviors. To help scientists understand systems better, an algorithm has been developed that estimates how similar the causes of one system’s behaviors are to the causes of another. This similarity between two systems is called their ”dynamical distance” from each other, and can be used to validate models, detect anomalies in a system, and explore how complex systems work. | en |
| dc.description.degree | Master of Science | en |
| dc.format.medium | ETD | en |
| dc.identifier.other | vt_gsexam:21401 | en |
| dc.identifier.uri | http://hdl.handle.net/10919/101659 | en |
| dc.publisher | Virginia Tech | en |
| dc.rights | In Copyright | en |
| dc.rights.uri | http://rightsstatements.org/vocab/InC/1.0/ | en |
| dc.subject | Data Analysis | en |
| dc.subject | Dynamical Kinds | en |
| dc.subject | Nonlinear Systems | en |
| dc.subject | Chaos | en |
| dc.subject | Automated Scientific Discovery | en |
| dc.subject | Order Identification | en |
| dc.title | Distinguishing Dynamical Kinds: An Approach for Automating Scientific Discovery | en |
| dc.type | Thesis | en |
| thesis.degree.discipline | Computer Science and Applications | en |
| thesis.degree.grantor | Virginia Polytechnic Institute and State University | en |
| thesis.degree.level | masters | en |
| thesis.degree.name | Master of Science | en |