The Experts below are selected from a list of 249 Experts worldwide ranked by ideXlab platform

Shaahin Angizi - One of the best experts on this subject based on the ideXlab platform.

  • Accelerating Low Bit-Width Deep Convolution Neural Network in MRAM
    2018 IEEE Computer Society Annual Symposium on VLSI (ISVLSI), 2018
    Co-Authors: Zhezhi He, Shaahin Angizi
    Abstract:

    Deep Convolution Neural Network (CNN) has achieved outstanding performance in image recognition over large scale dataset. However, pursuit of higher inference accuracy leads to CNN architecture with deeper layers and denser connections, which inevitably makes its hardware implementation demand more and more memory and computational resources. It can be interpreted as `CNN power and memory wall'. Recent research efforts have significantly reduced both model size and computational complexity by using low bit-width weights, activations and gradients, while keeping reasonably good accuracy. In this work, we present different emerging nonvolatile Magnetic Random Access Memory (MRAM) designs that could be leveraged to implement `bit-wise in-memory convolution engine', which could simultaneously Store Network parameters and compute low bit-width convolution. Such new computing model leverages the `in-memory computing' concept to accelerate CNN inference and reduce convolution energy consumption due to intrinsic logic-in-memory design and reduction of data communication.

  • ISVLSI - Accelerating Low Bit-Width Deep Convolution Neural Network in MRAM
    2018 IEEE Computer Society Annual Symposium on VLSI (ISVLSI), 2018
    Co-Authors: Zhezhi He, Shaahin Angizi
    Abstract:

    Deep Convolution Neural Network (CNN) has achieved outstanding performance in image recognition over large scale dataset. However, pursuit of higher inference accuracy leads to CNN architecture with deeper layers and denser connections, which inevitably makes its hardware implementation demand more and more memory and computational resources. It can be interpreted as ‘CNN power and memory wall’. Recent research efforts have significantly reduced both model size and computational complexity by using low bit-width weights, activations and gradients, while keeping reasonably good accuracy. In this work, we present different emerging nonvolatile Magnetic Random Access Memory (MRAM) designs that could be leveraged to implement ‘bit-wise in-memory convolution engine’, which could simultaneously Store Network parameters and compute low bit-width convolution. Such new computing model leverages the ‘in-memory computing’ concept to accelerate CNN inference and reduce convolution energy consumption due to intrinsic logic-in-memory design and reduction of data communication.

Gérard Cliquet - One of the best experts on this subject based on the ideXlab platform.

  • Retailers׳ expansion mode choice in foreign markets: Antecedents for expansion mode choice in the light of internationalization theories
    Journal of Retailing and Consumer Services, 2014
    Co-Authors: Karine Picot-coupey, Steve L Burt, Gérard Cliquet
    Abstract:

    Whenever a retail company expands its Store Network in a foreign market, decisions have to be made about how this can be achieved. Existing studies of retail internationalization have usually analyzed the "entry mode" choice as an end in itself, and not as the start of a firm's international development. In addition, there is much debate in the academic literature about the antecedents for retail foreign operation mode choice and the relevance of generic internationalization theories to international retailing. Therefore, the objectives of this research are (1) to investigate the paths of entry and subsequent expansion modes pursued by retailers in international markets and (2) to develop and test a model of expansion mode antecedents in the light of generic business internationalization theories. This is achieved on the basis of data collected from 43 French fashion retailers and a PLS-SEM approach. Results show that (1) retailers clearly differentiate between entry and expansion modes; (2) the international marketing plan, the perceived attractiveness of the foreign market, and strategic and ownership conditions are the key antecedents for the choice of an expansion mode. After comparing the results with the explanations proposed by the generic internationalization theories, a multi-theoretical framework is proposed which draws from the Uppsala internationalization process model, Network theory and the born-global theory. The findings provide a wealth of information for retailers' use in choosing appropriate foreign operation modes.

  • Retail internationalisation: explaining the expansion mode choice
    2007
    Co-Authors: Karine Picot-coupey, Gérard Cliquet
    Abstract:

    In order to develop their Store Network beyond their domestic market, retailers have to choose internationalisation modes not only to enter a market but also to expand in this market. Such decisions seem to be based on few certainties as they had enjoyed little interest in the literature so far. In this perspective, this research focuses on the determinants of Store Networks' expansion mode choice internationally.Based on an exploratory qualitative study, a conceptual model for determining the choice of an expansion mode in retail internationalisation is developed. Then, this model is tested using the PLS approach for path modeling

  • plural forms in Store Networks a model for Store Network evolution
    The International Review of Retail Distribution and Consumer Research, 2000
    Co-Authors: Gérard Cliquet
    Abstract:

    Very few papers have been written about plural forms in Store Networks. But today, many Store chains have both franchise and company-owned arrangements. Actually, this has been the case for more than twenty years. After a review of the literature devoted to the choice between franchise and company-owned systems, Bradach's research and his model of plural forms are described. This model was based on a study of five American fast food companies. Research conducted in France in the hotel and catering, bakery, and cosmetics industries, is then described. The advantages and drawbacks of plural forms are defined as they appear in the results of a survey among managers of twenty-one companies managing thirty-five chains. A model of the evolution of Store chain organization is then shown. This model takes into account various strategic and managerial considerations met by Store chains during their life cycle. Finally, chains and Networks are compared, and the results and research perspectives, discussed.

Zhezhi He - One of the best experts on this subject based on the ideXlab platform.

  • Accelerating Low Bit-Width Deep Convolution Neural Network in MRAM
    2018 IEEE Computer Society Annual Symposium on VLSI (ISVLSI), 2018
    Co-Authors: Zhezhi He, Shaahin Angizi
    Abstract:

    Deep Convolution Neural Network (CNN) has achieved outstanding performance in image recognition over large scale dataset. However, pursuit of higher inference accuracy leads to CNN architecture with deeper layers and denser connections, which inevitably makes its hardware implementation demand more and more memory and computational resources. It can be interpreted as `CNN power and memory wall'. Recent research efforts have significantly reduced both model size and computational complexity by using low bit-width weights, activations and gradients, while keeping reasonably good accuracy. In this work, we present different emerging nonvolatile Magnetic Random Access Memory (MRAM) designs that could be leveraged to implement `bit-wise in-memory convolution engine', which could simultaneously Store Network parameters and compute low bit-width convolution. Such new computing model leverages the `in-memory computing' concept to accelerate CNN inference and reduce convolution energy consumption due to intrinsic logic-in-memory design and reduction of data communication.

  • ISVLSI - Accelerating Low Bit-Width Deep Convolution Neural Network in MRAM
    2018 IEEE Computer Society Annual Symposium on VLSI (ISVLSI), 2018
    Co-Authors: Zhezhi He, Shaahin Angizi
    Abstract:

    Deep Convolution Neural Network (CNN) has achieved outstanding performance in image recognition over large scale dataset. However, pursuit of higher inference accuracy leads to CNN architecture with deeper layers and denser connections, which inevitably makes its hardware implementation demand more and more memory and computational resources. It can be interpreted as ‘CNN power and memory wall’. Recent research efforts have significantly reduced both model size and computational complexity by using low bit-width weights, activations and gradients, while keeping reasonably good accuracy. In this work, we present different emerging nonvolatile Magnetic Random Access Memory (MRAM) designs that could be leveraged to implement ‘bit-wise in-memory convolution engine’, which could simultaneously Store Network parameters and compute low bit-width convolution. Such new computing model leverages the ‘in-memory computing’ concept to accelerate CNN inference and reduce convolution energy consumption due to intrinsic logic-in-memory design and reduction of data communication.

Boon Thau Loo - One of the best experts on this subject based on the ideXlab platform.

  • Provenance-aware secure Networks
    Proceedings - International Conference on Data Engineering, 2008
    Co-Authors: Wenchao Zhou, Eric Cronin, Boon Thau Loo
    Abstract:

    Network accountability and forensic analysis have become increasingly important, as a means of performing Network diagnostics, identifying malicious nodes, enforcing trust management policies, and imposing diverse billing over the Internet. This has led to a series of work to provide better Network support for accountability, and efficient mechanisms to trace packets and information flows through the Internet. In this paper, we make the following contributions. First, we show that Network accountability and forensic analysis can be posed generally as data provenance computations and queries over distributed streams. In particular, one can utilize declarative Networks with appropriate security and provenance extensions to provide a unified declarative framework for specifying, analyzing and auditing Networks. Second, we propose a taxonomy of data provenance along multiple axes, and show that they map naturally to different use cases in Networks. Third, we suggest techniques to efficiently compute and Store Network provenance, and provide an initial performance evaluation on the P2 declarative Networking system with modifications to support authenticated communication and provenance.

Karine Picot-coupey - One of the best experts on this subject based on the ideXlab platform.

  • Determinants of retail Store Network expansion via shop-in-shops
    International Journal of Retail & Distribution Management, 2018
    Co-Authors: Karine Picot-coupey, Jean-laurent Viviani, Paul Amadieu
    Abstract:

    Why do some retail Networks operate shop-in-shops along with stand-alone units while others do not? Drawing on a resource-based and intellectual capital (IC) perspective as a broad theoretical lens, the purpose of this paper is to focus on retailer-run shop-in-shops and examine the determinants of their adoption.,To gain a comprehensive understanding of shop-in-shop adoption by retail branded Networks, a research design mixing a quantitative study (n = 170) and a qualitative study (n = 19) was adopted to test nine hypotheses regarding these determinants of the adoption of retailer-run shop-in-shops and explore in greater depth the processes whereby they actually occur.,The main findings show that intangible resources are major determinants of the choice to operate shop-in-shops while tangible resources are minor determinants. The more robust results of the analysis lie in the positive effect of own-label merchandise range, premium pricing strategy, positioning based on symbols, retail concept fast renewal and high sector specialisation on the choice to operate a shop-in-shop. The effect of financial constraints on the decision to expand via shop-in-shops is limited.,The authors emphasise the importance of marketing-related and company-related characteristics in differentiating the likelihood of retail Networks to expand via shop-in-shops. These results lend support to the relevance of a resource-based and IC perspective in explaining the propensity of retailers to develop via shop-in-shops.,The decision to operate shop-in-shops should depend on the extent to which intangible resources – the most important being retail positioning grounded in symbols, an own-label merchandise range, and a high retail branded Network reputation – can be valued and enhanced. Expanding a retail Network via shop-in-shops does not appear to be a financially constrained expansion strategy: it must be considered as a relevant first best strategy when an independent and young retail company has intangible resources to value but limited tangible resources.,The study contributes to channel management and retailing research in four ways. First, it precisely delineates the specific characteristics of shop-in-shops. Second, it provides theoretical explanations – based on a resource and IC perspective – of determinants that influence the choice of shop-in-shops. Third, it empirically tests the influence of marketing-related and company-related characteristics when adopting shop-in-shops. Fourth, it provides insights into how adopting shop-in-shops. To the authors’ knowledge, the research is on the first to analyse theoretically and test the determinants for the choice of retailer-run shop-in-shops.

  • Retailers׳ expansion mode choice in foreign markets: Antecedents for expansion mode choice in the light of internationalization theories
    Journal of Retailing and Consumer Services, 2014
    Co-Authors: Karine Picot-coupey, Steve L Burt, Gérard Cliquet
    Abstract:

    Whenever a retail company expands its Store Network in a foreign market, decisions have to be made about how this can be achieved. Existing studies of retail internationalization have usually analyzed the "entry mode" choice as an end in itself, and not as the start of a firm's international development. In addition, there is much debate in the academic literature about the antecedents for retail foreign operation mode choice and the relevance of generic internationalization theories to international retailing. Therefore, the objectives of this research are (1) to investigate the paths of entry and subsequent expansion modes pursued by retailers in international markets and (2) to develop and test a model of expansion mode antecedents in the light of generic business internationalization theories. This is achieved on the basis of data collected from 43 French fashion retailers and a PLS-SEM approach. Results show that (1) retailers clearly differentiate between entry and expansion modes; (2) the international marketing plan, the perceived attractiveness of the foreign market, and strategic and ownership conditions are the key antecedents for the choice of an expansion mode. After comparing the results with the explanations proposed by the generic internationalization theories, a multi-theoretical framework is proposed which draws from the Uppsala internationalization process model, Network theory and the born-global theory. The findings provide a wealth of information for retailers' use in choosing appropriate foreign operation modes.

  • Retail internationalisation: explaining the expansion mode choice
    2007
    Co-Authors: Karine Picot-coupey, Gérard Cliquet
    Abstract:

    In order to develop their Store Network beyond their domestic market, retailers have to choose internationalisation modes not only to enter a market but also to expand in this market. Such decisions seem to be based on few certainties as they had enjoyed little interest in the literature so far. In this perspective, this research focuses on the determinants of Store Networks' expansion mode choice internationally.Based on an exploratory qualitative study, a conceptual model for determining the choice of an expansion mode in retail internationalisation is developed. Then, this model is tested using the PLS approach for path modeling