教学课程资源库 更多>>

关键词： Computer science   Electronic devices  

摘要： In-memory computing using resistive memory devices is a promising non-von Neumann approach for making energy-efficient deep learning inference hardware. However, due to device variability and noise, the network needs to be trained in a specific way so that transferring the digitally trained weights to the analog resistive memory devices will not result in significant loss of accuracy. Here, we introduce a methodology to train ResNet-type convolutional neural networks that results in no appreciable accuracy loss when transferring weights to phase-change memory (PCM) devices. We also propose a compensation technique that exploits the batch normalization parameters to improve the accuracy retention over time. We achieve a classification accuracy of 93.7% on CIFAR-10 and a top-1 accuracy of 71.6% on ImageNet benchmarks after mapping the trained weights to PCM. Our hardware results on CIFAR-10 with ResNet-32 demonstrate an accuracy above 93.5% retained over a one-day period, where each of the 361,722 synaptic weights is programmed on just two PCM devices organized in a differential configuration. Designing deep learning inference hardware based on in-memory computing remains a challenge. Here, the authors propose a strategy to train ResNet-type convolutional neural networks which results in reduced accuracy loss when transferring weights to in-memory computing hardware based on phase-change memory.

关键词： Reference publication year spectroscopy (RPYS)   Citation analysis   Computer science   Eastern Europe   Web of science  

摘要： Purpose The current article presents the results of a case study dealing with the historical roots of Eastern European researchers in computer science. Design/methodology/approach The study is based on an analysis of cited references stemming from a collection of around 80,000 computer science papers by Eastern European researchers published from 1989 to 2014. By using a method called "reference publication year spectroscopy" (RPYS) for historical analyses based on bibliometric data, we analyze around 800,000 references cited in those papers. The study identifies the peak years, including most frequently cited publications (from 1952, 1965 and 1975), and focuses on these outstanding works for the field. The research shows how these influential papers were cited in Eastern Europe and in general, and on which scientific fields they have the most impact. Findings A noteworthy publication that seems to have a tremendous effect on Eastern European computer science is Zadeh's "Fuzzy sets" article which appeared in Information and Control in 1965. The study demonstrates that computer scientists from Eastern Europe are more conservative in their citation behaviour and tend to refer to older and more established research than their counterparts from the West. Originality/value Which are the historical roots of researchers working in a particular field or on a specific topic? Are there certain publications - landmark papers - which are important for their research? We guess that these are questions bothering researchers in many fields.

关键词： adappt intelligence   ADSK   Albert Gidari   artificial intelligence   artificial intelligence technologies   Autodesk   business   cafeterias   Charles Rechsteiner   computer science   consumer services   coronavirus   corporate   Dan Ryan   diversified holding companies   epidemics   general news   health   HON   Honeywell International   humanities   industrial electronics   industrial goods   industrial news   infectious diseases   Jamie Kinch   JLL   Jones Lang LaSalle   machine learning   medical conditions   novel coronaviruses   one-person elevators   outbreaks   plexiglass   political   Rajat Malhotra   rapid7   RDS.A   RDSA.LN   respiratory tract diseases   Royal Dutch Shell   RPD   sciences   security cameras   security systems   semiconductors   technology   Texas Instruments   tracking employee movements   TXN   vergesense   Vivek Bagri  

关键词： adappt intelligence   ADSK   Albert Gidari   artificial intelligence   artificial intelligence technologies   Autodesk   business   cafeterias   Charles Rechsteiner   computer science   consumer services   coronavirus   corporate   Dan Ryan   diversified holding companies   epidemics   general news   health   HON   Honeywell International   humanities   industrial electronics   industrial goods   industrial news   infectious diseases   Jamie Kinch   JLL   Jones Lang LaSalle   machine learning   medical conditions   novel coronaviruses   one-person elevators   outbreaks   plexiglass   political   Rajat Malhotra   rapid7   RDS.A   RDSA.LN   respiratory tract diseases   Royal Dutch Shell   RPD   sciences   security cameras   security systems   semiconductors   technology   Texas Instruments   tracking employee movements   TXN   vergesense   Vivek Bagri  

关键词： Computer science   Interdisciplinary instruction   Elementary   Problem-based learning   Implementation research  

摘要： Background As the number of computer science (CS) jobs become increasingly available in this country and computing skills become essential tools for managing all aspects of our personal lives, CS is quickly becoming an essential element of K-12 education and recently, there has been increased attention to bringing computer science to the elementary grades. However, with a schedule that emphasizes literacy and mathematics, and other subjects competing for instructional time, creating opportunities for CS in the elementary school day is challenging. This study aimed to address this problem by investigating the use of problem-based transdisciplinary modules (i.e., "Time4CS" modules) that combined English language arts (ELA), science, and social studies lessons with the "Fundamentals" CS education program. Results Results indicated that teachers who taught Time4CS modules completed more CS lessons than teachers who did not teach the modules. Further, across all classrooms, completing a higher percentage of non-grade level assigned Fundamentals lessons (i.e., lessons above or below grade level that were available to teachers, but not required for their particular grade level) was positively associated with students' achievement outcomes on state ELA and mathematics tests. Additionally, higher amounts of interdisciplinary teaching practices were associated with higher student achievement, specifically students' state assessment ELA scores. Conclusions This study demonstrated that transdisciplinary problem-based modules that integrate the teaching of CS with other subject areas are a feasible way to bring more CS opportunities to younger learners. Moreover, it showed that implementing such modules is linked to more positive student academic achievement outcomes. With attentive revision, the modules featured in this study may be useful tools for elementary schools. These findings have implications for researchers, school district administrators, and those individuals who ar

关键词： adappt intelligence   ADSK   Albert Gidari   artificial intelligence   artificial intelligence technologies   Autodesk   business   cafeterias   Charles Rechsteiner   computer science   consumer services   coronavirus   corporate   Dan Ryan   diversified holding companies   epidemics   general news   health   HON   Honeywell International   humanities   industrial electronics   industrial goods   industrial news   infectious diseases   Jamie Kinch   JLL   Jones Lang LaSalle   machine learning   medical conditions   novel coronaviruses   one-person elevators   outbreaks   plexiglass   political   Rajat Malhotra   rapid7   RDS.A   RDSA.LN   respiratory tract diseases   Royal Dutch Shell   RPD   sciences   security cameras   security systems   semiconductors   technology   Texas Instruments   tracking employee movements   TXN   vergesense   Vivek Bagri  

关键词： adappt intelligence   ADSK   Albert Gidari   artificial intelligence   artificial intelligence technologies   Autodesk   business   cafeterias   Charles Rechsteiner   computer science   consumer services   coronavirus   corporate   Dan Ryan   diversified holding companies   epidemics   general news   health   HON   Honeywell International   humanities   industrial electronics   industrial goods   industrial news   infectious diseases   Jamie Kinch   JLL   Jones Lang LaSalle   machine learning   medical conditions   novel coronaviruses   one-person elevators   outbreaks   plexiglass   political   Rajat Malhotra   rapid7   RDS.A   RDSA.LN   respiratory tract diseases   Royal Dutch Shell   RPD   sciences   security cameras   security systems   semiconductors   technology   Texas Instruments   tracking employee movements   TXN   vergesense   Vivek Bagri  

关键词： artificial intelligence   Beijing Infervision Technology   Bruno Beomonte Zobel   computer science   epidemics   Francesco De Cobelli   general news   health   Hugh Harvey   humanities   infectious diseases   Infervision (Beijing)   Ivan Blokhin   Lunit   machine learning   medical conditions   Moshe Becker   novel coronaviruses   outbreaks   political   Prashant Warier   Qure.ai   Qure.ai Technologies Pvt.   RADLogics   respiratory tract diseases   sciences   Sergey Morozov  

关键词： mercy   Religious Studies   mass incarceration   racism   Catholic social teaching   environmental destruction   systematic theology   capitalism   neoliberalism   migration  

ISBN: (数字)9780823290444 ISBN: (纸本)9780823288014

摘要： Contrary to Catholicism, Catholic social teaching, and the commitment to live out the mercy of Jesus Christ, today’s dominant global economic and cultural system, neoliberalism, demands that life be led as a series of sacrifices to the market. This book’s theological critique of neoliberalism begins with recent papal teaching against “economism,” proceeds into a historical and theoretical analysis of neoliberalism’s conception as a discourse in academia and the business community, its rise to global prominence through class warfare, its subtle redefining of human self-understanding via the notion of “human capital,” and its formation of an ethos of mercilessness. Central is treatment of four neoliberal-perpetuated and -exacerbated crises: environmental destruction, slum proliferation, mass incarceration, and mass deportation. This entails plumbing the sacrificial and racist depths of neoliberalism. The book offers an antineoliberal systematic theology founded on Trinitarian mercy, a neighbor anthropology and innkeeper ecclesiology, and a politics of mercy, or a civilizational program grounded in, yet reimagining, the traditional Catholic works of mercy. This coheres with a “playbook” for social transformation that uses the universal destination of goods and abolitionism to direct the corporal works of mercy against the neoliberal utopianism that brought enhanced ecological devastation, slum growth, mass imprisonment, and abuse of migrants. In concert with official Catholic teaching, the Gospel injunction to “be merciful,” and hopeful visions of various people of good will, Send Lazarus urges a robust antineoliberal and antiracist politics, which amounts to a joyous expression of Christic hope for abundant life.

关键词： artificial intelligence   Beijing Infervision Technology   Bruno Beomonte Zobel   computer science   epidemics   Francesco De Cobelli   general news   health   Hugh Harvey   humanities   infectious diseases   Infervision (Beijing)   Ivan Blokhin   Lunit   machine learning   medical conditions   Moshe Becker   novel coronaviruses   outbreaks   political   Prashant Warier   Qure.ai   Qure.ai Technologies Pvt.   RADLogics   respiratory tract diseases   sciences   Sergey Morozov