5 Hits in 5.2 sec

gLOP: the global and Local Penalty for Capturing Predictive Heterogeneity [article]

Rhiannon V. Rose, Daniel J. Lizotte
2016 arXiv   pre-print
We present gLOP, the global and Local Penalty, a framework for capturing predictive heterogeneity and identifying predictive outliers. gLOP is based on penalized regression for multitask learning, which  ...  We give two optimization algorithms for gLOP, one space-efficient, and another giving the full regularization path.  ...  On the degrees of freedom of the lasso. The Annals of Statistics, 35(5):2173-2192, 2007. gLOP: the global and Local Penalty for Capturing Predictive Heterogeneity  ... 
arXiv:1608.00027v1 fatcat:bjmhv5bx6fas3e7orzu6nkxl7q


Yu Feng, Yuhao Zhu
2019 Proceedings of the 46th International Symposium on Computer Architecture - ISCA '19  
The reactive strategy leads to local optimizations that schedule event executions one at a time, missing global optimization opportunities. This paper proposes Proactive Event Scheduling (PES).  ...  The key idea of PES is to proactively anticipate future events and thereby globally coordinate scheduling decisions across events.  ...  Acknowledgement We thank Brian Anderson, Ben Hayden, and Eric Seckler from the Google Chrome team and anonymous discussions from the Chromium-dev platform.  ... 
doi:10.1145/3307650.3322248 dblp:conf/isca/FengZ19 fatcat:ri26ptgupfewjicr7zaewhgz34

Designing, Modeling, and Optimizing Data-Intensive Computing Systems [article]

Gagandeep Singh
2022 arXiv   pre-print
Third, we propose NAPEL, an ML-based application performance and energy prediction framework for data-centric architectures.  ...  First, we design NERO, a data-centric accelerator for a real-world weather prediction application.  ...  Acknowledgements We hope that the ideas, analyses, methods, and techniques presented in this dissertation will enable the development of energy-efficient data-intensive computing systems and drive the  ... 
arXiv:2208.08886v1 fatcat:ledjwayl6jdbfmgp2i45vooore

GPU accelerated onboard data processing for downlink optimisation

Rebecca Davidson
The processing throughput, compression ratio and power consumption of the new CCSDS-123 image compression GPU application are assessed and characterised for a desktop GPU and the onboard representative  ...  This data bottleneck must be alleviated in order for EO satellites to efficiently deliver the quality and quantity of payload data now expected by its reliant applications.  ...  Possible Values: [Full or Reduced] Prediction Mode The local differences used for prediction.  ... 
doi:10.15126/thesis.00852470 fatcat:napelwugefei7pmh3txxtg4rr4


(:Unkn) Unknown, University, My, Justin Y. Shi
For that, I agreed to take on the feasibility study of SMC for practical applications.  ...  An overlay network for distributed and parallel computing application, such as Blockchain, has been proven to defy the "common wisdom" for two important distributed computing challenges: a) Extreme scale  ...  It runs the SMC protocol on top of TCP/IP protocols. • Gateway is responsible for the local and global storage content addressable networks.  ... 
doi:10.34944/dspace/909 fatcat:ijz4yxsairgivfd5bwlwuwjuiq