331 Hits in 0.71 sec

Recovery Guarantee of Non-negative Matrix Factorization via Alternating Updates [article]

Yuanzhi Li, Yingyu Liang, Andrej Risteski
2016 arXiv   pre-print
Non-negative matrix factorization is a popular tool for decomposing data into feature and weight matrices under non-negativity constraints. It enjoys practical success but is poorly understood theoretically. This paper proposes an algorithm that alternates between decoding the weights and updating the features, and shows that assuming a generative model of the data, it provably recovers the ground-truth under fairly mild conditions. In particular, its only essential requirement on features is
more » ... near independence. Furthermore, the algorithm uses ReLU to exploit the non-negativity for decoding the weights, and thus can tolerate adversarial noise that can potentially be as large as the signal, and can tolerate unbiased noise much larger than the signal. The analysis relies on a carefully designed coupling between two potential functions, which we believe is of independent interest.
arXiv:1611.03819v1 fatcat:qy3ib7lbfnewhl56a4vw7jg74u

Provable Alternating Gradient Descent for Non-negative Matrix Factorization with Strong Correlations [article]

Yuanzhi Li, Yingyu Liang
2017 arXiv   pre-print
This work was done when Yingyu Liang was visiting the Simons Institute.  ...  Yuanzhi Li <>, Yingyu Liang <>. Proceedings of the 34 th International Conference on Machine Learning, Sydney, Australia, PMLR 70, 2017.  ... 
arXiv:1706.04097v1 fatcat:ri75f64psnhblh4mbfu3kg65ea

A Concise Biogenetically Inspired Formal Synthesis of Camptothecin

Yuanzhi Li, Mengqian Zhu, Liang Xu
2021 Youji huaxue  
A concise biogenetically inspired formal synthesis of camptothecin without use of protecting groups has been developed starting from tryptamine and easily prepared ethyl glutaconaldehyde salt. The synthesis features the key Pictet-Spengler reaction, efficient intramolecular oxa-Diels-Alder reaction to construct heptacyclic monoterpenoid indole alkaloid intermediate, as well as the following Winterfeldt biomimetic oxidation to form quinolinone moiety from indole skeleton.
doi:10.6023/cjoc202102034 fatcat:4hvjp77ksvgvdm7emrhvamhy4u

Nanoscale triboelectrification gated transistor

Tianzhao Bu, Liang Xu, Zhiwei Yang, Xiang Yang, Guoxu Liu, Yuanzhi Cao, Chi Zhang, Zhong Lin Wang
2020 Nature Communications  
Tribotronics has attracted great attention owing to the demonstrated triboelectrification-controlled electronics and established direct modulation mechanism by external mechanical stimuli. Here, a nanoscale triboelectrification-gated transistor has been studied with contact-mode atomic force microscopy and scanning Kevin probe microscopy. The detailed working principle was analyzed at first, in which the nanoscale triboelectrification can tune the carrier transport in the transistor. Then with
more » ... he manipulated nanoscale triboelectrification, the effects of contact force, scan speed, contact cycles, contact region and charge diffusion on the transistor were investigated, respectively. Moreover, the manipulated nanoscale triboelectrification serving as a rewritable floating gate has demonstrated different modulation effects by an applied tip voltage. This work has realized the nanoscale triboelectric modulation on electronics, which could provide a deep understanding for the theoretical mechanism of tribotronics and may have great applications in nanoscale transistor, micro/nano-electronic circuit and nano-electromechanical system.
doi:10.1038/s41467-020-14909-6 pmid:32103025 fatcat:kuis7qiqgrhd5epduguvb4mp3a

VrR-VG: Refocusing Visually-Relevant Relationships [article]

Yuanzhi Liang, Yalong Bai, Wei Zhang, Xueming Qian, Li Zhu, Tao Mei
2019 arXiv   pre-print
Relationships encode the interactions among individual instances, and play a critical role in deep visual scene understanding. Suffering from the high predictability with non-visual information, existing methods tend to fit the statistical bias rather than "learning" to "infer" the relationships from images. To encourage further development in visual relationships, we propose a novel method to automatically mine more valuable relationships by pruning visually-irrelevant ones. We construct a new
more » ... scene-graph dataset named Visually-Relevant Relationships Dataset (VrR-VG) based on Visual Genome. Compared with existing datasets, the performance gap between learnable and statistical method is more significant in VrR-VG, and frequency-based analysis does not work anymore. Moreover, we propose to learn a relationship-aware representation by jointly considering instances, attributes and relationships. By applying the representation-aware feature learned on VrR-VG, the performances of image captioning and visual question answering are systematically improved with a large margin, which demonstrates the gain of our dataset and the features embedding schema. VrR-VG is available via
arXiv:1902.00313v2 fatcat:d43grgaz6zdexeafkam5bjzfsm

Observation of Mollow triplet with metastability exchange collisions in 3He atoms [article]

Yuanzhi Zhan, Xiang Peng, Sheng Li, Liang Zhang, Jingbiao Chen, and Hong Guo
2018 arXiv   pre-print
We study the dressed states of 3He atoms and experimentally observe the Mollow triplet (MT) induced with an ultra-low-frequency (ULF) oscillating magnetic field as low as 4 Hz. The ULF MT signatures from the ground states of 3He atoms are transferred to the metastable states by metastability-exchange collisions (MECs) and measured optically, which demonstrates 2 s coherence time in the dressed ground states. The result shows the possibility of ULF magnetic field amplitude measurement and a new scheme for optical frequency modulation.
arXiv:1810.11779v1 fatcat:ounoir6j65cwbnww4chkj3r6fi

Recovery guarantee of weighted low-rank approximation via alternating minimization [article]

Yuanzhi Li, Yingyu Liang, Andrej Risteski
2016 arXiv   pre-print
Many applications require recovering a ground truth low-rank matrix from noisy observations of the entries, which in practice is typically formulated as a weighted low-rank approximation problem and solved by non-convex optimization heuristics such as alternating minimization. In this paper, we provide provable recovery guarantee of weighted low-rank via a simple alternating minimization algorithm. In particular, for a natural class of matrices and weights and without any assumption on the
more » ... , we bound the spectral norm of the difference between the recovered matrix and the ground truth, by the spectral norm of the weighted noise plus an additive error that decreases exponentially with the number of rounds of alternating minimization, from either initialization by SVD or, more importantly, random initialization. These provide the first theoretical results for weighted low-rank via alternating minimization with non-binary deterministic weights, significantly generalizing those for matrix completion, the special case with binary weights, since our assumptions are similar or weaker than those made in existing works. Furthermore, this is achieved by a very simple algorithm that improves the vanilla alternating minimization with a simple clipping step. The key technical challenge is that under non-binary deterministic weights, na\"ive alternating steps will destroy the incoherence and spectral properties of the intermediate solutions, which are needed for making progress towards the ground truth. We show that the properties only need to hold in an average sense and can be achieved by the clipping step. We further provide an alternating algorithm that uses a whitening step that keeps the properties via SDP and Rademacher rounding and thus requires weaker assumptions. This technique can potentially be applied in some other applications and is of independent interest.
arXiv:1602.02262v2 fatcat:d4c3jx6s4fejhl5xvpuvhtvnhe

Comparison of Oil Spill Classifications Using Fully and Compact Polarimetric SAR Images

Yuanzhi Zhang, Yu Li, X. Liang, Jinyeu Tsou
2017 Applied Sciences  
In this paper, we present a comparison between several algorithms for oil spill classifications using fully and compact polarimetric SAR images. Oil spill is considered as one of the most significant sources of marine pollution. As a major difficulty of SAR-based oil spill detection algorithms is the classification between mineral and biogenic oil, we focus on quantitatively analyzing and comparing fully and compact polarimetric satellite synthetic aperture radar (SAR) modes to detect
more » ... n slicks over the sea surface, discriminating them from weak-damping surfactants, such as biogenic slicks. The experiment was conducted on quad-pol SAR data acquired during the Norwegian oil-on-water experiment in 2011. A universal procedure was used to extract the features from quad-, dual-and compact polarimetric SAR modes to rank different polarimetric SAR modes and common supervised classifiers. Among all the dual-and compact polarimetric SAR modes, the π/2 mode has the best performance. The best supervised classifiers vary and depended on whether sufficient polarimetric information can be obtained in each polarimetric mode. We also analyzed the influence of the number of polarimetric parameters considered as inputs for the supervised classifiers, onto the detection/discrimination performance. We discovered that a feature set with four features is sufficient for most polarimetric feature-based oil spill classifications. Moreover, dimension reduction algorithms, including principle component analysis (PCA) and the local linear embedding (LLE) algorithm, were employed to learn low dimensional and distinctive information from quad-polarimetric SAR features. The performance of the new feature sets has comparable performance in oil spill classification. These accidents and illegal acts cause damage to the coastal ecosystem, emphasizing the importance of detecting oil spills in their early stages. Although remote sensing with optical sensors can be used in oil spill detection, they are unavoidably restricted by weather and light conditions. Therefore, satellite synthetic aperture radar (SAR) data from ERS-1/2 (European Remote Sensing Satellites), ENVISAT (Environmental Satellite), ALOS (Advanced Land Observing Satellite), RADARSAT-1/2 and TerraSAR-X have been widely used to detect and monitor oil spills [1] [2] [3] [4] [5] [6] [7] [8] due to the large spatial coverage, all-weather conditions and imaging capability during day-night times [9] . In addition, airborne SAR sensors, such as Uninhabited Aerial Vehicle Synthetic Aperture Radar (UAVSAR) developed by JPL at L-band and E-SAR (developed by DLR), have proven their potential for scientific research on ocean or land [10, 11] . In SAR images, the detection of oil slick on the sea surface relies on the detection/quantification of its attenuation of Bragg scattering on the sea surface. When Bragg scattering happens, the signals from different sea surface facets interfere with each other. Moreover, according to the composite sea surface model, the roughness of the sea surface can be seen as small-scale capillary waves (contributing to Bragg scattering) superimposed on large-scale gravity waves. An illustration of this model can be seen in Figure 1 . The sea surface of the oil-covered region appears smoother than its surrounding area. This is because the Bragg scattering of these areas is suppressed by the presence of hydrocarbons. However, the main backscatter from the sea surface is contributed by Bragg scattering. As a result, in SAR images the oil slick-covered area can usually be detected as a very dark (low backscattered) area.
doi:10.3390/app7020193 fatcat:f343ok7bsfhmrak7pu2zqhhf7m

Experimental Study on Diesel Engine Emission Characteristics Based on Different Exhaust Pipe Coating Schemes

Keqin Zhao, Diming Lou, Yunhua Zhang, Liang Fang, Yuanzhi Tang
2021 Micromachines  
The thermal insulation performance of exhaust pipes coated with various materials (basalt and glass fiber materials) under different braiding forms (sleeve, winding and felt types) and the effects on the emission characteristics of diesel engines were experimentally studied through engine bench tests. The results indicated that the thermal insulation performance of basalt fiber was higher than that of glass fiber, and more notably advantageous at the early stage of the diesel engine idle cold
more » ... ase. The average temperature drop during the first 600 s of the basalt felt (BF) pipe was 2.6 °C smaller than that of the glass fiber felt (GF) pipe. Comparing the different braiding forms, the temperature decrease in the felt-type braided material was 2.6 °C and 2.9 °C smaller than that in the sleeve- and winding-type braided materials, respectively. The basalt material was better than the glass fiber material regarding the gaseous pollutant emission reduction performance, especially in the idling cold phase of diesel engines. The NOx conversion rate of the BF pipe was 7.4% higher than that of the GF pipe, and the hydrocarbon (HC) conversion rate was 2.3% higher than that of the GF pipe, while the CO conversion rate during the first 100 s was 24.5% higher than that of the GF pipe. However, the particulate matter emissions were not notably different.
doi:10.3390/mi12101155 pmid:34683206 fatcat:hcqfg2f4bbauzm4exjyfgutr3y

Copper Nanowires as Fully Transparent Conductive Electrodes

Huizhang Guo, Na Lin, Yuanzhi Chen, Zhenwei Wang, Qingshui Xie, Tongchang Zheng, Na Gao, Shuping Li, Junyong Kang, Duanjun Cai, Dong-Liang Peng
2013 Scientific Reports  
In pondering of new promising transparent conductors to replace the cost rising tin-doped indium oxide (ITO), metal nanowires have been widely concerned. Herein, we demonstrate an approach for successful synthesis of long and fine Cu nanowires (NWs) through a novel catalytic scheme involving nickel ions. Such Cu NWs in high aspect ratio (diameter of 16.2 6 2 nm and length up to 40 mm) provide long distance for electron transport and, meanwhile, large space for light transmission. Transparent
more » ... ctrodes fabricated using the Cu NW ink achieve a low sheet resistance of 1.4 Ohm/sq at 14% transmittance and a high transparency of 93.1% at 51.5 Ohm/sq. The flexibility and stability were tested with 100-timebending by 1806and no resistance change occurred. Ohmic contact was achieved to the pand n-GaN on blue light emitting diode chip and bright electroluminescence from the front face confirmed the excellent transparency.
doi:10.1038/srep02323 pmid:23900572 pmcid:PMC3728602 fatcat:vzhvogm6jncevnhlh7g5gmvbjy

Observation of Mollow quintuplet in F=3/2 hyperfine structure state of 3He atomic cell [article]

Yuanzhi Zhan, Xiang Peng, Sheng Li, Liang Zhang, Jingbiao Chen, and Hong Guo
2018 arXiv   pre-print
We experimentally observed the Mollow quintuplet (MQ) in F=3/2 hyperfine structure state of 3He atoms. The metastability-exchange collisions (MECs) transfer the Mollow Triplet (MT) from the ground states of 3He atoms to the metastable states, and the MQ is demonstrated by four Zeeman levels of F=3/2 hyperfine states with linearly polarized light. The similar effect also achieves in the mixture cell of 3He and 4He.
arXiv:1812.07992v1 fatcat:iaz76abzxng5pd3llmncb2j3tq

Learning Mixtures of Linear Regressions with Nearly Optimal Complexity [article]

Yuanzhi Li, Yingyu Liang
2020 arXiv   pre-print
Acknowledgements Yingyu Liang would like to acknowledge that support for this research was provided by the Office of the Vice Chancellor for Research and Graduate Education at the University of Wisconsin  ...  Another line of works used tensor methods for MLR [Chaganty and Liang, 2013, Sedghi et al., 2016] .  ...  ., [Chaganty and Liang, 2013, Klusowski et al., 2017] ) or designing new algorithms (e.g., [Chen et al., 2014] ).  ... 
arXiv:1802.07895v3 fatcat:yy3k5jccz5fa5en5q5zbqub4ei

Effect of Water on the Dynamic Tensile Mechanical Properties of Calcium Silicate Hydrate: Based on Molecular Dynamics Simulation

Jikai Zhou, Yuanzhi Liang
2019 Materials  
To study the effect of water on the dynamic mechanical properties of calcium silicate hydrate (C–S–H) at the atomic scale, the molecular dynamics simulations were performed in uniaxial tension with different strain rates for C–S–H with a degree of saturation from 0% to 100%. Our calculations demonstrate that the dynamic tensile mechanical properties of C–S–H decrease with increasing water content and increase with increasing strain rates. With an increase in the degree of saturation, the strain
more » ... rate sensitivity of C–S–H tends to increase. According to Morse potential function, the tensile stress-strain relationship curves of C–S–H are decomposed and fitted, and the dynamic tensile constitutive relationship of C–S–H considering the effect of water content is proposed. This reveals the strain rate effect of the cementitious materials with different water content from molecular insights, and the dynamic constitutive relationship obtained in this paper is necessary to the modelling of cementitious materials at the meso-scale.
doi:10.3390/ma12172837 pmid:31484393 pmcid:PMC6747960 fatcat:nvtkughrpza2dcuhheruu25suu

A Nonaqueous Approach to the Preparation of Iron Phosphide Nanowires

Houde She, Yuanzhi Chen, Ruitao Wen, Kui Zhang, Guang-Hui Yue, Dong-Liang Peng
2010 Nanoscale Research Letters  
Previous preparation of iron phosphide nanowires usually employed toxic and unstable iron carbonyl compounds as precursor. In this study, we demonstrate that iron phosphide nanowires can be synthesized via a facile nonaqueous chemical route that utilizes a commonly available iron precursor, iron (III) acetylacetonate. In the synthesis, trioctylphosphine (TOP) and trioctylphosphine oxide (TOPO) have been used as surfactants, and oleylamine has been used as solvent. The crystalline structure and
more » ... orphology of the as-synthesized products were characterized by powder X-ray diffraction (XRD) and transmission electron microscopy (TEM). The obtained iron phosphide nanowires have a typical width of *16 nm and a length of several hundred nanometers. Structural and compositional characterization reveals a hexagonal Fe 2 P crystalline phase. The morphology of as-synthesized products is greatly influenced by the ratio of TOP/TOPO. The presence of TOPO has been found to be essential for the growth of high-quality iron phosphide nanowires. Magnetic measurements reveal ferromagnetic characteristics, and hysteresis behaviors below the blocking temperature have been observed.
doi:10.1007/s11671-010-9559-4 pmid:20672078 pmcid:PMC2894222 fatcat:a3wucxvno5e65ff4jqm5lf2ouy

Preparation and Characterization of Hydroxyapatite Coating on AZ31 Magnesium Alloy Induced by Carboxymethyl Cellulose-Dopamine

Yanxia Yang, Yuanzhi Wu, Yu Wei, Tian Zeng, Baocheng Cao, Jun Liang
2021 Materials  
(Yuanzhi Wu), Y.W. (Yu Wei) and T.Z.; data curation, Y.Y. and Y.W. (Yuanzhi Wu); writing-original draft preparation, Y.Y., Y.W. (Yuanzhi Wu); writing-review and editing, Y.W.  ... 
doi:10.3390/ma14081849 pmid:33917900 fatcat:6xqoqcnl3jb4hoxwe6g6eexacq
« Previous Showing results 1 — 15 out of 331 results