Battery lifetime prediction at early cycles is crucial for researchers and manufacturers to examine product quality and promote technology *** learning has been widely utilized to construct data-driven solutions for h...
详细信息
Battery lifetime prediction at early cycles is crucial for researchers and manufacturers to examine product quality and promote technology *** learning has been widely utilized to construct data-driven solutions for high-accuracy ***,the internal mechanisms of batteries are sensitive to many factors,such as charging/discharging protocols,manufacturing/storage conditions,and usage *** factors will induce state transitions,thereby decreasing the prediction accuracy of data-driven *** learning is a promising technique that overcomes this difficulty and achieves accurate predictions by jointly utilizing information from various ***,we develop two transfer learning methods,Bayesian Model Fusion and Weighted Orthogonal Matching Pursuit,to strategically combine prior knowledge with limited information from the target dataset to achieve superior prediction *** our results,our transfer learning methods reduce root-mean-squared error by 41%through adapting to the target ***,the transfer learning strategies identify the variations of impactful features across different sets of batteries and therefore disentangle the battery degradation mechanisms and the root cause of state transitions from the perspective of data *** findings suggest that the transfer learning strategies proposed in our work are capable of acquiring knowledge across multiple data sources for solving specialized issues.
Healthcare information systems are complex and critical enterprise systems that link together geographically distributed hospitals, clinics, physician offices and other business units with distinct business functions ...
详细信息
Healthcare information systems are complex and critical enterprise systems that link together geographically distributed hospitals, clinics, physician offices and other business units with distinct business functions and mutual dependencies. In the past, these systems were built based on proprietary solutions, acquired in piecemeal fashion and tightly coupled through ad hoc means. This resulted in stovepipe systems that had many duplicated functions and that were monolithic, non-extensible and non-interoperable. How to migrate from these stovepipe systems to the next-generation open healthcare information systems that are interoperable, extensible and maintainable is increasingly a pressing problem for the healthcare industry. In this paper, we present our experience in addressing the problem using an architecture-centered approach for enterprise system development and integration based on the distributed object technology standards OMA/CORBA. Our case study is conducted from a user organization (as opposed to developer organization) point of view and is based on a large-scale effort undertaken at the Baptist Health systems of South Florida, a large healthcare organization serving the South Florida region. Key lessons learnt from this case study include: (1) establishing a clear architectural vision is essential to successful enterprise system development; (2) adoption of a standard architecture and infrastructure is the best approach to achieve interoperable, extensible and cohesive enterprise systems; and (3) an effective development methodology and business process are key to implementing the architectural vision.
暂无评论