As the smart grid becomes reality, software architectures for integrating legacy systems with new innovative approaches for grid management are needed. These architectures must exhibit flexibility, extensibility, inte...
详细信息
High Resolution Medical Images provide a useful visual and analytic tool to medical professionals. These images often require substantial disk space and thus may provide bottlenecks in I/O access and network transmiss...
详细信息
ISBN:
(纸本)9780889869202
High Resolution Medical Images provide a useful visual and analytic tool to medical professionals. These images often require substantial disk space and thus may provide bottlenecks in I/O access and network transmission, and create a challenge for disk storage capacity. These problems can be mitigated by using compression algorithms that not only reduce the storage requirements of the data files, but can do so quickly as to meet real time workloads. We present a lossless entropy based compression scheme that achieves both of these goals for medical image data. We demonstrate the effectiveness of our lossless algorithm using a medical image dataset and compare its performance with other lossless compression techniques. Our method provides competitive compression ratios without sacrificing real-time performance.
As the smart grid becomes reality, software architectures for integrating legacy systems with new innovative approaches for grid management are needed. These architectures must exhibit flexibility, extensibility, inte...
详细信息
As the smart grid becomes reality, software architectures for integrating legacy systems with new innovative approaches for grid management are needed. These architectures must exhibit flexibility, extensibility, interoperability and scalability. In this position paper, we describe our preliminary work to design such an architecture, known as GridOPTICS, that will enable the deployment and integration of new software tools in smart grid operations. Our preliminary design is based upon use cases from PNNL's Future Power Grid Initiative, which is developing a collection of advanced software technologies for smart grid management and control. We describe the motivations for GridOPTICS, and the preliminary design that we are currently prototyping for several distinct use cases.
In this paper, we investigate the use of an encoder-decoder neural architecture for unsupervised representation learning of mixed asynchronous data, and we introduce the JMETTS (Joint Modelling of Event Traces and Tim...
详细信息
In this paper, we investigate the use of an encoder-decoder neural architecture for unsupervised representation learning of mixed asynchronous data, and we introduce the JMETTS (Joint Modelling of Event Traces and Time Series) model. Our aim is to perform short-term forecasts for multivariate time series contextualized by events. As a proof of concept, we focus on a real-world case related to digitally assisted training in anaesthesiology. With a maximal prediction error percentage around 5.5%, the high predictive performance of JMETTS is comparable to its only competitor published to date. The source code is publicly available.
暂无评论