GSTDTAP

浏览/检索结果: 共24条,第1-10条 帮助

已选(0)清除 条数/页:   排序方式:
Development of advanced artificial intelligence models for daily rainfall prediction 期刊论文
ATMOSPHERIC RESEARCH, 2020, 237
作者:  Binh Thai Pham;  Lu Minh Le;  Tien-Thinh Le;  Kien-Trinh Thi Bui;  Vuong Minh Le;  Hai-Bang Ly;  Prakash, Indra
收藏  |  浏览/下载:12/0  |  提交时间:2020/07/02
Rainfall  Artificial Neural Networks  Robustness analysis  Support Vector Machines  Adaptive Network based Fuzzy Inference System  Particle Swarm Optimization  
Physics-Informed Deep Neural Networks for Learning Parameters and Constitutive Relationships in Subsurface Flow Problems 期刊论文
WATER RESOURCES RESEARCH, 2020, 56 (5)
作者:  Tartakovsky, A. M.;  Marrero, C. Ortiz;  Perdikaris, Paris;  Tartakovsky, G. D.;  Barajas-Solano, D.
收藏  |  浏览/下载:12/0  |  提交时间:2020/07/02
deep neural networks  physics-informed machine learning  parameter estimation  learning constitutive relationships  unsaturated flow  MAP  
A map of object space in primate inferotemporal cortex 期刊论文
NATURE, 2020, 583 (7814) : 103-+
作者:  Wu, Huihui;  Li, Bosheng;  Iwakawa, Hiro-oki;  Pan, Yajie;  Tang, Xianli;  Ling-hu, Qianyan;  Liu, Yuelin;  Sheng, Shixin;  Feng, Li;  Zhang, Hong;  Zhang, Xinyan;  Tang, Zhonghua;  Xia, Xinli;  Zhai, Jixian;  Guo, Hongwei
收藏  |  浏览/下载:47/0  |  提交时间:2020/07/03

Primate inferotemporal cortex contains a coarse map of object space consisting of four networks, identified using functional imaging, electrophysiology and deep networks.


The inferotemporal (IT) cortex is responsible for object recognition, but it is unclear how the representation of visual objects is organized in this part of the brain. Areas that are selective for categories such as faces, bodies, and scenes have been found(1-5), but large parts of IT cortex lack any known specialization, raising the question of what general principle governs IT organization. Here we used functional MRI, microstimulation, electrophysiology, and deep networks to investigate the organization of macaque IT cortex. We built a low-dimensional object space to describe general objects using a feedforward deep neural network trained on object classification(6). Responses of IT cells to a large set of objects revealed that single IT cells project incoming objects onto specific axes of this space. Anatomically, cells were clustered into four networks according to the first two components of their preferred axes, forming a map of object space. This map was repeated across three hierarchical stages of increasing view invariance, and cells that comprised these maps collectively harboured sufficient coding capacity to approximately reconstruct objects. These results provide a unified picture of IT organization in which category-selective regions are part of a coarse map of object space whose dimensions can be extracted from a deep network.


  
Probabilistic Forecasting of El Nino Using Neural Network Models 期刊论文
GEOPHYSICAL RESEARCH LETTERS, 2020, 47 (6)
作者:  Petersik, Paul Johannes;  Dijkstra, Henk A.
收藏  |  浏览/下载:7/0  |  提交时间:2020/07/02
El Nino  prediction  machine learning  neural networks  probabilistic forecasting  
A 450 million years long latitudinal gradient in age-dependent extinction 期刊论文
ECOLOGY LETTERS, 2020, 23 (3) : 439-446
作者:  Silvestro, Daniele;  Castiglione, Silvia;  Mondanaro, Alessandro;  Serio, Carmela;  Melchionna, Marina;  Piras, Paolo;  Di Febbraro, Mirko;  Carotenuto, Francesco;  Rook, Lorenzo;  Raia, Pasquale
收藏  |  浏览/下载:7/0  |  提交时间:2020/07/02
'  law of constant extinction'  deep learning  fossil occurrences  neural networks  mass extinction  
Neuronal programming by microbiota regulates intestinal physiology 期刊论文
NATURE, 2020, 578 (7794) : 284-+
作者:  Li, Yilong;  Roberts, Nicola D.;  Wala, Jeremiah A.;  Shapira, Ofer;  Schumacher, Steven E.;  Kumar, Kiran;  Khurana, Ekta;  Waszak, Sebastian;  Korbel, Jan O.;  Haber, James E.;  Imielinski, Marcin;  Weischenfeldt, Joachim;  Beroukhim, Rameen;  Campbell, Peter J.;  Akdemir, Kadir C.;  Alvarez, Eva G.;  Baez-Ortega, Adrian;  Boutros, Paul C.;  Bowtell, David D. L.;  Brors, Benedikt;  Burns, Kathleen H.;  Chan, Kin;  Chen, Ken;  Cortes-Ciriano, Isidro;  Dueso-Barroso, Ana;  Dunford, Andrew J.;  Edwards, Paul A.;  Estivill, Xavier;  Etemadmoghadam, Dariush;  Feuerbach, Lars;  Fink, J. Lynn;  Frenkel-Morgenstern, Milana;  Garsed, Dale W.;  Gerstein, Mark;  Gordenin, Dmitry A.;  Haan, David;  Hess, Julian M.;  Hutter, Barbara;  Jones, David T. W.;  Ju, Young Seok;  Kazanov, Marat D.;  Klimczak, Leszek J.;  Koh, Youngil;  Lee, Eunjung Alice;  Lee, Jake June-Koo;  Lynch, Andy G.;  Macintyre, Geoff;  Markowetz, Florian;  Martincorena, Inigo;  Martinez-Fundichely, Alexander;  Meyerson, Matthew;  Miyano, Satoru;  Nakagawa, Hidewaki;  Navarro, Fabio C. P.;  Ossowski, Stephan;  Park, Peter J.;  Pearson, John, V;  Puiggros, Montserrat;  Rippe, Karsten;  Roberts, Steven A.;  Rodriguez-Martin, Bernardo;  Scully, Ralph;  Shackleton, Mark;  Sidiropoulos, Nikos;  Sieverling, Lina;  Stewart, Chip;  Torrents, David;  Tubio, Jose M. C.;  Villasante, Izar;  Waddell, Nicola;  Yang, Lixing;  Yao, Xiaotong;  Yoon, Sung-Soo;  Zamora, Jorge;  Zhang, Cheng-Zhong
收藏  |  浏览/下载:40/0  |  提交时间:2020/07/03

Neural control of the function of visceral organs is essential for homeostasis and health. Intestinal peristalsis is critical for digestive physiology and host defence, and is often dysregulated in gastrointestinal disorders(1). Luminal factors, such as diet and microbiota, regulate neurogenic programs of gut motility(2-5), but the underlying molecular mechanisms remain unclear. Here we show that the transcription factor aryl hydrocarbon receptor (AHR) functions as a biosensor in intestinal neural circuits, linking their functional output to the microbial environment of the gut lumen. Using nuclear RNA sequencing of mouse enteric neurons that represent distinct intestinal segments and microbiota states, we demonstrate that the intrinsic neural networks of the colon exhibit unique transcriptional profiles that are controlled by the combined effects of host genetic programs and microbial colonization. Microbiota-induced expression of AHR in neurons of the distal gastrointestinal tract enables these neurons to respond to the luminal environment and to induce expression of neuron-specific effector mechanisms. Neuron-specific deletion of Ahr, or constitutive overexpression of its negative feedback regulator CYP1A1, results in reduced peristaltic activity of the colon, similar to that observed in microbiota-depleted mice. Finally, expression of Ahr in the enteric neurons of mice treated with antibiotics partially restores intestinal motility. Together, our experiments identify AHR signalling in enteric neurons as a regulatory node that integrates the luminal environment with the physiological output of intestinal neural circuits to maintain gut homeostasis and health.


In a mouse model, aryl hydrocarbon receptor signalling in enteric neurons is revealed as a mechanism that helps to maintain gut homeostasis by integrating the luminal environment with the physiology of intestinal neural circuits.


  
Recurrent interactions in local cortical circuits 期刊论文
NATURE, 2020, 579 (7798) : 256-+
作者:  Liu, Yang;  Nguyen, Phong T.;  Wang, Xun;  Zhao, Yuting;  Meacham, Corbin E.;  Zou, Zhongju;  Bordieanu, Bogdan;  Johanns, Manuel;  Vertommen, Didier;  Wijshake, Tobias;  May, Herman;  Xiao, Guanghua;  Shoji-Kawata, Sanae;  Rider, Mark H.
收藏  |  浏览/下载:8/0  |  提交时间:2020/07/03

Most cortical synapses are local and excitatory. Local recurrent circuits could implement amplification, allowing pattern completion and other computations(1-4). Cortical circuits contain subnetworks that consist of neurons with similar receptive fields and increased connectivity relative to the network average(5,6). Cortical neurons that encode different types of information are spatially intermingled and distributed over large brain volumes(5-7), and this complexity has hindered attempts to probe the function of these subnetworks by perturbing them individually(8). Here we use computational modelling, optical recordings and manipulations to probe the function of recurrent coupling in layer 2/3 of the mouse vibrissal somatosensory cortex during active tactile discrimination. A neural circuit model of layer 2/3 revealed that recurrent excitation enhances sensory signals by amplification, but only for subnetworks with increased connectivity. Model networks with high amplification were sensitive to damage: loss of a few members of the subnetwork degraded stimulus encoding. We tested this prediction by mapping neuronal selectivity(7) and photoablating(9,10) neurons with specific selectivity. Ablation of a small proportion of layer 2/3 neurons (10-20, less than 5% of the total) representing touch markedly reduced responses in the spared touch representation, but not in other representations. Ablations most strongly affected neurons with stimulus responses that were similar to those of the ablated population, which is also consistent with network models. Recurrence among cortical neurons with similar selectivity therefore drives input-specific amplification during behaviour.


Computational modelling, imaging and single-cell ablation in layer 2/3 of the mouse vibrissal somatosensory cortex reveals that recurrent activity in cortical neurons can drive input-specific amplification during behaviour.


  
Classification with a disordered dopantatom network in silicon 期刊论文
NATURE, 2020, 577 (7790) : 341-+
作者:  Vagnozzi, Ronald J.;  Maillet, Marjorie;  Sargent, Michelle A.;  Khalil, Hadi;  Johansen, Anne Katrine Z.;  Schwanekamp, Jennifer A.;  York, Allen J.;  Huang, Vincent;  Nahrendorf, Matthias;  Sadayappan, Sakthivel;  Molkentin, Jeffery D.
收藏  |  浏览/下载:24/0  |  提交时间:2020/07/03

Classification is an important task at which both biological and artificial neural networks excel(1,2). In machine learning, nonlinear projection into a high-dimensional feature space can make data linearly separable(3,4), simplifying the classification of complex features. Such nonlinear projections are computationally expensive in conventional computers. A promising approach is to exploit physical materials systems that perform this nonlinear projection intrinsically, because of their high computational density(5), inherent parallelism and energy efficiency(6,7). However, existing approaches either rely on the systems'  time dynamics, which requires sequential data processing and therefore hinders parallel computation(5,6,8), or employ large materials systems that are difficult to scale up(7). Here we use a parallel, nanoscale approach inspired by filters in the brain(1) and artificial neural networks(2) to perform nonlinear classification and feature extraction. We exploit the nonlinearity of hopping conduction(9-11) through an electrically tunable network of boron dopant atoms in silicon, reconfiguring the network through artificial evolution to realize different computational functions. We first solve the canonical two-input binary classification problem, realizing all Boolean logic gates(12) up to room temperature, demonstrating nonlinear classification with the nanomaterial system. We then evolve our dopant network to realize feature filters(2) that can perform four-input binary classification on the Modified National Institute of Standards and Technology handwritten digit database. Implementation of our material-based filters substantially improves the classification accuracy over that of a linear classifier directly applied to the original data(13). Our results establish a paradigm of silicon-based electronics for smallfootprint and energy-efficient computation(14).


  
Fully hardware-implemented memristor convolutional neural network 期刊论文
NATURE, 2020, 577 (7792) : 641-+
作者:  Yoshioka-Kobayashi, Kumiko;  Matsumiya, Marina;  Niino, Yusuke;  Isomura, Akihiro;  Kori, Hiroshi;  Miyawaki, Atsushi;  Kageyama, Ryoichiro
收藏  |  浏览/下载:39/0  |  提交时间:2020/07/03

Memristor-enabled neuromorphic computing systems provide a fast and energy-efficient approach to training neural networks(1-4). However, convolutional neural networks (CNNs)-one of the most important models for image recognition(5)-have not yet been fully hardware-implemented using memristor crossbars, which are cross-point arrays with a memristor device at each intersection. Moreover, achieving software-comparable results is highly challenging owing to the poor yield, large variation and other non-ideal characteristics of devices(6-9). Here we report the fabrication of high-yield, high-performance and uniform memristor crossbar arrays for the implementation of CNNs, which integrate eight 2,048-cell memristor arrays to improve parallel-computing efficiency. In addition, we propose an effective hybrid-training method to adapt to device imperfections and improve the overall system performance. We built a five-layer memristor-based CNN to perform MNIST10 image recognition, and achieved a high accuracy of more than 96 per cent. In addition to parallel convolutions using different kernels with shared inputs, replication of multiple identical kernels in memristor arrays was demonstrated for processing different inputs in parallel. The memristor-based CNN neuromorphic system has an energy efficiency more than two orders of magnitude greater than that of state-of-the-art graphics-processing units, and is shown to be scalable to larger networks, such as residual neural networks. Our results are expected to enable a viable memristor-based non-von Neumann hardware solution for deep neural networks and edge computing.


  
Detecting Climate Change Effects on Vb Cyclones in a 50-Member Single-Model Ensemble Using Machine Learning 期刊论文
GEOPHYSICAL RESEARCH LETTERS, 2019
作者:  Mittermeier, M.;  Braun, M.;  Hofstaetter, M.;  Wang, Y.;  Ludwig, R.
收藏  |  浏览/下载:13/0  |  提交时间:2020/02/17
Vb-cyclones  Machine Learning  Artificial Neural Networks (ANN)  Single-Model Large Ensembles  Internal Variability  Floods