1,721,015 research outputs found
Contrastive Dual Gating: Learning Sparse Features With Contrastive Learning
No full textContrastive learning (or its variants) has recently become a promising direction in the self-supervised learning domain, achieving similar performance as supervised learning with minimum fine-tuning. Despite the labeling efficiency, wide and large networks are required to achieve high accuracy, which incurs a high amount of computation and hinders the pragmatic merit of self-supervised learning. To effectively reduce the computation of insignificant features or channels, recent dynamic pruning algorithms for supervised learning employed auxiliary salience predictors. However, we found that such salience predictors cannot be easily trained when they are naïvely applied to contrastive learning from scratch. To address this issue, we propose contrastive dual gating (CDG), a novel dynamic pruning algorithm that skips the uninformative features during contrastive learning without hurting the trainability of the networks. We demonstrate the superiority of CDG with ResNet models for CIFAR-10, CIFAR-100, and ImageNet-100 datasets. Compared to our implementations of state-of-the-art dynamic pruning algorithms for self-supervised learning, CDG achieves up to 15% accuracy improvement for CIFAR-10 dataset with higher computation reduction
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Variations on the Author
Full text link“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
Appropriate Similarity Measures for Author Cocitation Analysis
Full text linkWe provide a number of new insights into the methodological discussion about author cocitation analysis. We first argue that the use of the Pearson correlation for measuring the similarity between authors’ cocitation profiles is not very satisfactory. We then discuss what kind of similarity measures may be used as an alternative to the Pearson correlation. We consider three similarity measures in particular. One is the well-known cosine. The other two similarity measures have not been used before in the bibliometric literature. Finally, we show by means of an example that our findings have a high practical relevance.information science;Pearson correlation;cosine;similarity measure;author cocitation analysis
Dispelling the Myths Behind First-author Citation Counts
Full text linkWe conducted a full-scale evaluative citation analysis study of scholars in the XML research field to explore just how different from each other author rankings resulting from different citation counting methods actually are, and to demonstrate the capability of emerging data and tools on the Web in supporting more realistic citation counting methods. Our results contest some common arguments for the continued
use of first-author citation counts in the evaluation of scholars, such as high correlations between author rankings by first-author citation counts and other citation
counting methods, and high costs of using more realistic citation counting methods that are not well-supported by the ISI databases. It is argued that increasingly available digital full text research papers make it possible for citation analysis studies to go beyond what the ISI databases have directly supported and to employ more
sophisticated methods
koamabayili/VECTRON-author-checklist: VECTRON author checklist
No full textWe have done our best to complete the author checklist relating to the use of animals in the hut study. Note that the objective for the hut study was to evaluate the IRS treatment applications for residual efficacy against Anopheles mosquitoes, including the local An. coluzzii mosquito population. Cows were only used to attract mosquitoes into the huts and no tests were carried out directly on the cows. The author checklist is intended for use with studies where experiments are carried out on animals, which is why we have had such difficulty in completing this for the hut study, as many of the questions do not relate to how the cows were used
Efficient In- and Near- Memory Computing System With Hardware and Software Co-Design
Full text linkOver the past decades, we have entered the big data era, where increasing amounts of data are generated daily, requiring processing and analysis by computing systems. However, modern computing systems suffer from power-hungry data transfer and processing due to the separate design of storage and processing units. This separation leads to a significant gap between supply and demand when processing large amounts of data, especially on resource-limited devices such as IoT (Internet of Things) devices. Motivated by this, this dissertation focuses on software and hardware co-design across different layers (device, circuit, architecture, algorithm) to develop energy-efficient and high-performance in-/near-memory computing systems, accelerating data-intensive applications such as deep learning and bioinformatics. This dissertation explores how designing memory and logic through innovations in devices, circuits, and architecture can overcome existing memory and power walls. Additionally, hardware-aware algorithm design further optimizes hardware utilization, significantly enhancing the computing efficiency of modern non-Von Neumann systems.
This dissertation begins by introducing circuit and algorithm optimizations to enhance the energy efficiency of RRAM-based in-memory computing architectures for on-device multi-task adaptation in neural networks. Here, the utilization of Resistive RAM (RRAM) takes the form of a 1 transistor 1 resistor (1T1R) crossbar array, enabling the integration of dot-product operations within the memory itself. This approach not only minimizes energy consumption during memory access but also boosts the throughput and energy efficiency of dot-product operations. The column-wise mask-based method has been proposed to avoid the high reprogramming energy consumption of RRAM for on-device multi-task learning. Through hardware-aware design innovation, the necessary masking operation to adapt to a new task can be seamlessly implemented in existing crossbar-based convolution engines with minimal hardware and memory overhead. More importantly, this approach significantly reduces power-hungry cell reprogramming while providing a trade-off between accuracy and energy consumption.
The next focus of this dissertation involves machine learning accelerator designs utilizing hybrid memory technologies for on-device continual learning. By integrating processing elements (PEs) made with different memory types (such as SRAM, RRAM, and MRAM) and distributing operations among these varied PEs, the design effectively mitigates the drawbacks associated with using a single-memory PE, such as high write energy, latency, and instability. Moreover, the hybrid design not only reduces area and power consumption but also maintains high accuracy, offering a scalable and versatile solution for on-device continual learning.
Finally, the dissertation presents designs for other data-intensive applications such as Min/Max searching and accelerating genome processing tasks, leveraging both volatile and non-volatile memories. For genome processing, an RRAM-based macro prototype is fabricated using the monolithic integration of HfO RRAM and 65-nm CMOS technology, achieving remarkable energy efficiency metrics of 2.07 TOPS/W (Tera-Operations Per Second per Watt) and 2.12G suffixes/Joule at 1.0V. This represents the most energy-efficient solution to date for genome processing tasks
- …
