Developing and Implementing the Data Mining Algorithms in RAVEN

Abstract

The RAVEN code is becoming a comprehensive tool to perform probabilistic risk assessment, uncertainty quantification, and verification and validation. The RAVEN code is being developed to support many programs and to provide a set of methodologies and algorithms for the analysis of data sets, simulation of physical phenomenon, etc. Scientific computer codes can generate enormous amounts of data. To post-process and analyze such data might, in some cases, take longer than the initial software runtime. Data mining algorithms/methods help in recognizing and understanding patterns in the data, and thus discover knowledge in databases.The methodologies used in dynamic probabilistic risk assessment or in uncertainty and error quantification analysis couple system/physics codes with simulation controller codes, such as RAVEN. RAVEN introduces both deterministic and stochastic elements into the simulation while the system/physics code model the dynamics deterministically. A typical analysis is performed by sampling values of a set of parameter attributes. A major challenge in using dynamic probabilistic risk assessment or uncertainty and error quantification analysis for a complex system is to analyze the large number of scenarios generated. Data mining techniques are typically used to better organize and understand data, i.e. recognizing patterns in the data. This report focuses on development and implementation of a post processing infrastructure and Application Programming Interfaces (APIs) for different data mining algorithms within RAVEN to extend its capability. The application of these algorithms to different databases will be also presented.

Enhancements to the RAVEN code in FY16

“…Automated techniques systematically partition the domain, typically in a greedy fashion, by splitting the data along either an original axis-aligned dimension or a reduced axis after performing dimensionality reduction. The criterion for splitting the data varies from minimizing numerical error [2,10,13,17,27] to more geometry and topology-based criteria [29,38].…”

“…hybrid dynamic event tree and adaptive hybrid dynamic even tree) [1], advanced static data mining capabilities (e.g. clustering, principal component analysis, manifold learning) [2], and ways to connect multiple reduced order model (able to reproduce scalar figure of merits) in order to create ensemble of models.…”

Data Mining Usage and Applications in Health Services

“…The answer as seen in Fig. 2 is that these areas, having association with KDD, focuses on all the processes of exploration of verbose information, how to store and retrieve the data, how algorithms can be scaled to huge data sets and still work effectively [23,24], how the results can be interpreted and visualized, and how the whole humanmachine interaction can be usefully modeled and supported.…”