Filters








28 Hits in 5.7 sec

Enabling Flexibility for Sparse Tensor Acceleration via Heterogeneity [article]

Eric Qin, Raveesh Garg, Abhimanyu Bambhaniya, Michael Pellauer, Angshuman Parashar, Sivasankaran Rajamanickam, Cong Hao, Tushar Krishna
<span title="2022-01-21">2022</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
A common characteristic among all of these accelerators is that they target tensor algebra (typically matrix multiplications); yet dozens of new accelerators are proposed for every new application.  ...  AESPA, a heterogeneous sparse accelerator design template constructed with the sub-accelerators generated from HARD TACO, and (3) a suite of scheduling strategies to map tensor kernels onto heterogeneous  ...  Hard TACO takes a unique approach by utilizing the output of an established sparse tensor compiler [8] to generate RTL.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2201.08916v1">arXiv:2201.08916v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/z2u3rn6qgzewlpwkhnputccw7a">fatcat:z2u3rn6qgzewlpwkhnputccw7a</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20220129054410/https://arxiv.org/pdf/2201.08916v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/56/07/56075a802608422b79a245c6c6f87882220824a7.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2201.08916v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Sparseloop: An Analytical Approach To Sparse Tensor Accelerator Modeling [article]

Yannan Nellie Wu, Po-An Tsai, Angshuman Parashar, Vivienne Sze, Joel S. Emer
<span title="2022-05-12">2022</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
In recent years, many accelerators have been proposed to efficiently process sparse tensor algebra applications (e.g., sparse neural networks).  ...  The lack of systematic description and modeling support for these sparse tensor accelerators impedes hardware designers from efficient and effective design space exploration.  ...  ACKNOWLEDGMENTS We thank Haoquan Zhang for discussions on statistical analysis of tensor density. We would also like to thank the anonymous reviewers for their constructive feedback.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2205.05826v1">arXiv:2205.05826v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/sywqrcvxh5hzdgen6p545sxl6e">fatcat:sywqrcvxh5hzdgen6p545sxl6e</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20220514051029/https://arxiv.org/pdf/2205.05826v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/57/8c/578cd61fca7a4e9c735d3e16ae2d89c9cc2faf80.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2205.05826v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

IntersectX: An Efficient Accelerator for Graph Mining [article]

Gengyu Rao, Jingji Chen, Jason Yik, Xuehai Qian
<span title="2021-04-19">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
In this paper, we propose IntersectX, a vertically designed accelerator for pattern enumeration with stream instruction set extension and architectural supports based on conventional processor.  ...  value computations; and (5) the nested intersection translator that generates micro-op sequences for implementing nested intersections.  ...  : an architecture for the general sparse tensor computation or specifically targeting DNN acceleration.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2012.10848v4">arXiv:2012.10848v4</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/fpsomsej7zhbpaxsepdgksmzu4">fatcat:fpsomsej7zhbpaxsepdgksmzu4</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210303071733/https://arxiv.org/pdf/2012.10848v3.pdf" title="fulltext PDF download [not primary version]" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <span style="color: #f43e3e;">&#10033;</span> <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/d3/6a/d36a03fd4dfdc69709a5e547adb7f0b9315cd3e5.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2012.10848v4" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Extending Sparse Tensor Accelerators to Support Multiple Compression Formats [article]

Eric Qin, Geonhwa Jeong, William Won, Sheng-Chun Kao, Hyoukjun Kwon, Sudarshan Srinivasan, Dipankar Das, Gordon E. Moon, Sivasankaran Rajamanickam, Tushar Krishna
<span title="2021-03-18">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
This work proposes hardware extensions to accelerators for supporting numerous format combinations seamlessly and demonstrates ~4X speedup over performing format conversions in software.  ...  We demonstrate that both the compactness of different compression formats and the compute efficiency of the algorithms enabled by them vary across tensor dimensions and amount of sparsity.  ...  We also thank the anonymous reviewers for their valuable feedback. Support for this work was provided through the ARIAA co-design center funded by the U.S.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2103.10452v1">arXiv:2103.10452v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/jsn7psgnhra4zngrwt7hhgdzs4">fatcat:jsn7psgnhra4zngrwt7hhgdzs4</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210323223422/https://arxiv.org/ftp/arxiv/papers/2103/2103.10452.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/d9/0b/d90bcdcb6e7e091108cd8bacd7bb0bd8b1aa1a0a.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2103.10452v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Capstan: A Vector RDA for Sparsity [article]

Alexander Rucker, Matthew Vilim, Tian Zhao, Yaqi Zhang, Raghu Prabhakar, Kunle Olukotun
<span title="2021-04-26">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
This paper proposes Capstan: a scalable, parallel-patterns-based, reconfigurable-dataflow accelerator (RDA) for sparse and dense tensor applications.  ...  For a variety of sparse applications, Capstan with DDR4 memory is 22x faster than a multi-core CPU baseline, while Capstan with HBM2 memory is 17x faster than an Nvidia V100 GPU.  ...  RELATED WORK Graph & Sparse Linear Algebra Accelerators.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2104.12760v1">arXiv:2104.12760v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/k7s6dsgikvgixcip2xyrd7eriu">fatcat:k7s6dsgikvgixcip2xyrd7eriu</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210429143653/https://arxiv.org/pdf/2104.12760v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/2f/96/2f962ef9b2109d8ca88fcc7e74d472f4ac843f16.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2104.12760v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Accelerating Sparse DNN Models without Hardware-Support via Tile-Wise Sparsity [article]

Cong Guo and Bo Yang Hsueh and Jingwen Leng and Yuxian Qiu and Yue Guan and Zehuan Wang and Xiaoying Jia and Xipeng Li and Minyi Guo and Yuhao Zhu
<span title="2020-08-29">2020</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Consequently, sparse models cannot achieve meaningful speedup on commodity hardware (e.g., GPU) built for dense matrix computations.  ...  We implement and evaluate the sparsity pattern on GPU tensor core, achieving a 1.95x speedup over the dense model.  ...  ACKNOWLEDGEMENT We thank the anonymous reviewers for their constructive feedback for improving the work.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2008.13006v1">arXiv:2008.13006v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/5r3luayqivaafbt5daho4rulmi">fatcat:5r3luayqivaafbt5daho4rulmi</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200902152209/https://arxiv.org/pdf/2008.13006v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2008.13006v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Barrier-Free Large-Scale Sparse Tensor Accelerator (BARISTA) For Convolutional Neural Networks [article]

Ashish Gondimalla, Sree Charan Gundabolu, T.N. Vijaykumar, Mithuna Thottethodi
<span title="2021-05-08">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
To that end, we propose the barrier-free large-scale sparse tensor accelerator (BARISTA).  ...  BARISTA (1) is the first architecture for scaling up sparse CNN accelerators; (2) reduces on-chip bandwidth demand by telescoping request-combining the input map requests and snarfing the filter requests  ...  ExTensor [25] proposes hierarchical representations for sparse tensors.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2104.08734v2">arXiv:2104.08734v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/i3l5er5vebhpnpj3kbh4cadzmm">fatcat:i3l5er5vebhpnpj3kbh4cadzmm</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210512181601/https://arxiv.org/pdf/2104.08734v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/f0/fd/f0fdd7c52549d9e71bb45b9535005653ea59ac44.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2104.08734v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Hardware Acceleration of Sparse and Irregular Tensor Computations of ML Models: A Survey and Insights [article]

Shail Dave, Riyadh Baghdadi, Tony Nowatzki, Sasikanth Avancha, Aviral Shrivastava, Baoxin Li
<span title="2021-07-22">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
The takeaways from this paper include: understanding the key challenges in accelerating sparse, irregular-shaped, and quantized tensors; understanding enhancements in accelerator systems for supporting  ...  structured sparsity can improve storage efficiency and balance computations; understanding how to compile and map models with sparse tensors on the accelerators; understanding recent design trends for  ...  ENCODINGS FOR COMPRESSING SPARSE TENSORS A sparse tensor is compressed with an encoding format. An encoded tensor contains actual data (NZ values) and metadata (information about positions of NZs).  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2007.00864v2">arXiv:2007.00864v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/k4o2xboh4vbudadfiriiwjp7uu">fatcat:k4o2xboh4vbudadfiriiwjp7uu</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210724091537/https://arxiv.org/pdf/2007.00864v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/73/03/730353ede6bf5a1a478fe2e44d34b9436ca48a1c.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2007.00864v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Sextans: A Streaming Accelerator for General-Purpose Sparse-Matrix Dense-Matrix Multiplication [article]

Linghao Song, Yuze Chi, Atefeh Sohrabizadeh, Young-kyu Choi, Jason Lau, Jason Cong
<span title="2022-01-13">2022</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
In this paper, we present Sextans, an accelerator for general-purpose SpMM processing.  ...  Sextans accelerator features (1) fast random access using on-chip memory, (2) streaming access to off-chip large matrices, (3) PE-aware non-zero scheduling for balanced workload with an II=1 pipeline,  ...  Some accelerators also target for high order tensor operation such as metricized tensor times Khatri Rao product (MTTKRP) and tensor times matrix chain (TTMc).  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2109.11081v2">arXiv:2109.11081v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/wlf7lraenzd7tkb3g7h6mumqei">fatcat:wlf7lraenzd7tkb3g7h6mumqei</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210925100012/https://arxiv.org/pdf/2109.11081v1.pdf" title="fulltext PDF download [not primary version]" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <span style="color: #f43e3e;">&#10033;</span> <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/e6/9e/e69e9dafb65bd1b6ca61567b748af72eb1bbb2e2.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2109.11081v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Synergistic CPU-FPGA Acceleration of Sparse Linear Algebra [article]

Mohammadreza Soltaniyeh, Richard P. Martin, Santosh Nagarakatte
<span title="2020-04-29">2020</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
This paper describes REAP, a software-hardware approach that enables high performance sparse linear algebra computations on a cooperative CPU-FPGA platform.  ...  REAP improves the performance of Sparse General Matrix Multiplication (SpGEMM) and Sparse Cholesky Factorization by 3.2X and 1.85X compared to widely used sparse libraries for them on the CPU, respectively  ...  However, TPU is designed for dense matrices. Similarly, Extensor [55] is an ASIC that supports high-dimensional sparse data known as tensors and helps to match the non-zero elements quickly.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2004.13907v1">arXiv:2004.13907v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/ced5hi2yfvgg3lyau2xrdrtqjq">fatcat:ced5hi2yfvgg3lyau2xrdrtqjq</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200501232146/https://arxiv.org/pdf/2004.13907v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2004.13907v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

SPOTS: An Accelerator for Sparse Convolutional Networks Leveraging Systolic General Matrix-Matrix Multiplication [article]

Mohammadreza Soltaniyeh, Richard P. Martin, Santosh Nagarakatte
<span title="2021-11-24">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
This paper proposes a new hardware accelerator for sparse convolutional neural networks (CNNs) by building a hardware unit to perform the Image to Column (IM2COL) transformation of the input feature map  ...  Further, our design improves performance by effectively mapping the sparse data to the hardware units by utilizing sparsity in both input feature maps and weights.  ...  Tensor Processing Unit (TPU) [25] is an ASIC that has matrix multiplication as its core computation block to accelerate CNNs.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2107.13386v2">arXiv:2107.13386v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/k7oampka5rdztojmmwrr2yvnfm">fatcat:k7oampka5rdztojmmwrr2yvnfm</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20211201161532/https://arxiv.org/pdf/2107.13386v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/10/d0/10d024ac0a3dd7b6e669762da22913789976d354.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2107.13386v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Copernicus: Characterizing the Performance Implications of Compression Formats Used in Sparse Workloads [article]

Bahar Asgari, Ramyad Hadidi, Joshua Dierberger, Charlotte Steinichen, Amaan Marfatia, Hyesoon Kim
<span title="2021-10-18">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
To fill this gap of knowledge, we characterize the impact of using seven frequently used sparse formats on performance, based on a DSA for sparse matrix-vector multiplication (SpMV), implemented on an  ...  The primary challenge with sparse matrices has been efficiently storing and transferring data, for which many sparse formats have been proposed to significantly eliminate zero entries.  ...  architecture for processing sparse matrices and an HLS-based implementation on an FPGA that can be used as a building block in further accelerators for sparse problems.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2011.10932v2">arXiv:2011.10932v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/dkc77jbaujailkbtepsmedvebq">fatcat:dkc77jbaujailkbtepsmedvebq</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20211020214429/https://arxiv.org/pdf/2011.10932v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/ac/fa/acfa743c4e56dda6d481f1c09dce63fd9b115925.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2011.10932v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Proximal Reinforcement Learning: A New Theory of Sequential Decision Making in Primal-Dual Spaces [article]

Sridhar Mahadevan, Bo Liu, Philip Thomas, Will Dabney, Steve Giguere, Nicholas Jacek, Ian Gemp, Ji Liu
<span title="2014-05-26">2014</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
This key technical innovation makes it possible to finally design "true" stochastic gradient methods for reinforcement learning.  ...  The Legendre transform elegantly generalizes past algorithms for solving reinforcement learning problems, such as natural gradient methods, which we show relate closely to the previously unconnected framework  ...  Principal funding for this research was provided by the National Science Foundation under the grant NSF IIS-1216467.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1405.6757v1">arXiv:1405.6757v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/u77kqc6iyncy7fixlnrfcnqrmy">fatcat:u77kqc6iyncy7fixlnrfcnqrmy</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200922005939/https://arxiv.org/pdf/1405.6757v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/fc/d0/fcd02a03977b3a70d16861a0d995fe6a91fc2b2c.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1405.6757v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Mathematics of Human Motion: from Animation towards Simulation (A View form the Outside) [article]

A.I. Zhmakin
<span title="2011-02-24">2011</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
On the other hand, he happens to be a human (who occasionally is moving) and, as everybody else, rates himself an expert in Applied Common Sense.  ...  Thus the author hopes that this view from the outside will be of some interest not only for the strangers like himself, but for those who are inside as well.  ...  It is claimed in the cited paper, that geometric algebra is computationally superior in comparison with the classical vector and tensor notations.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1102.4992v1">arXiv:1102.4992v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/vf63c7dtqfdsra562sxlb7uh2m">fatcat:vf63c7dtqfdsra562sxlb7uh2m</a> </span>
<a target="_blank" rel="noopener" href="https://archive.org/download/arxiv-1102.4992/1102.4992.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> File Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/d8/c6/d8c60e57bc117c0a2a7b4ac6ac56e4b63d79a184.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1102.4992v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Muscle Synergies Facilitate Computational Prediction of Subject-Specific Walking Motions

Andrew J. Meyer, Ilan Eskinazi, Jennifer N. Jackson, Anil V. Rao, Carolynn Patten, Benjamin J. Fregly
<span title="2016-10-13">2016</span> <i title="Frontiers Media SA"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/tkuhrcyiufdxtkdmjqvay6f2ua" style="color: black;">Frontiers in Bioengineering and Biotechnology</a> </i> &nbsp;
an individual post-stroke.  ...  However, there is currently no objective means for prescribing and implementing treatments that are likely to maximize recovery of walking function for any particular patient.  ...  We also eliminated nine muscle-tendon actuators without related EMG data (extensor hallucis longus, flexor hallucis longus, gemelli, gracilis, pectineus, piriformis, quadratus femoris, sartorius, tensor  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.3389/fbioe.2016.00077">doi:10.3389/fbioe.2016.00077</a> <a target="_blank" rel="external noopener" href="https://www.ncbi.nlm.nih.gov/pubmed/27790612">pmid:27790612</a> <a target="_blank" rel="external noopener" href="https://pubmed.ncbi.nlm.nih.gov/PMC5061852/">pmcid:PMC5061852</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/pm4jdpfgarcu5gizi33y7kioge">fatcat:pm4jdpfgarcu5gizi33y7kioge</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20201103043929/https://escholarship.org/content/qt84h3r42d/qt84h3r42d.pdf?t=qajztc" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/2b/a7/2ba714242cefb4ef836249d112c7118cf9f251b1.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.3389/fbioe.2016.00077"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="unlock alternate icon" style="background-color: #fb971f;"></i> frontiersin.org </button> </a> <a target="_blank" rel="external noopener" href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5061852" title="pubmed link"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> pubmed.gov </button> </a>
&laquo; Previous Showing results 1 &mdash; 15 out of 28 results