Showing papers by "University of Science and Technology of China published in 2014"

Black phosphorus field-effect transistors

Abstract: Two-dimensional crystals have emerged as a class of materials that may impact future electronic technologies. Experimentally identifying and characterizing new functional two-dimensional materials is challenging, but also potentially rewarding. Here, we fabricate field-effect transistors based on few-layer black phosphorus crystals with thickness down to a few nanometres. Reliable transistor performance is achieved at room temperature in samples thinner than 7.5 nm, with drain current modulation on the order of 10(5) and well-developed current saturation in the I-V characteristics. The charge-carrier mobility is found to be thickness-dependent, with the highest values up to ∼ 1,000 cm(2) V(-1) s(-1) obtained for a thickness of ∼ 10 nm. Our results demonstrate the potential of black phosphorus thin crystals as a new two-dimensional material for applications in nanoelectronic devices.

Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition

Abstract: Existing deep convolutional neural networks (CNNs) require a fixed-size (e.g., 224x224) input image. This requirement is "artificial" and may reduce the recognition accuracy for the images or sub-images of an arbitrary size/scale. In this work, we equip the networks with another pooling strategy, "spatial pyramid pooling", to eliminate the above requirement. The new network structure, called SPP-net, can generate a fixed-length representation regardless of image size/scale. Pyramid pooling is also robust to object deformations. With these advantages, SPP-net should in general improve all CNN-based image classification methods. On the ImageNet 2012 dataset, we demonstrate that SPP-net boosts the accuracy of a variety of CNN architectures despite their different designs. On the Pascal VOC 2007 and Caltech101 datasets, SPP-net achieves state-of-the-art classification results using a single full-image representation and no fine-tuning. The power of SPP-net is also significant in object detection. Using SPP-net, we compute the feature maps from the entire image only once, and then pool features in arbitrary regions (sub-images) to generate fixed-length representations for training the detectors. This method avoids repeatedly computing the convolutional features. In processing test images, our method is 24-102x faster than the R-CNN method, while achieving better or comparable accuracy on Pascal VOC 2007. In ImageNet Large Scale Visual Recognition Challenge (ILSVRC) 2014, our methods rank #2 in object detection and #3 in image classification among all 38 teams. This manuscript also introduces the improvement made for this competition.

Precise and Ultrafast Molecular Sieving Through Graphene Oxide Membranes

Abstract: Graphene-based materials can have well-defined nanometer pores and can exhibit low frictional water flow inside them, making their properties of interest for filtration and separation. We investigate permeation through micrometer-thick laminates prepared by means of vacuum filtration of graphene oxide suspensions. The laminates are vacuum-tight in the dry state but, if immersed in water, act as molecular sieves, blocking all solutes with hydrated radii larger than 4.5 angstroms. Smaller ions permeate through the membranes at rates thousands of times faster than what is expected for simple diffusion. We believe that this behavior is caused by a network of nanocapillaries that open up in the hydrated state and accept only species that fit in. The anomalously fast permeation is attributed to a capillary-like high pressure acting on ions inside graphene capillaries.

Mixed transition-metal oxides: design, synthesis, and energy-related applications.

Abstract: A promising family of mixed transition-metal oxides (MTMOs) (designated as Ax B3-x O4 ; A, B=Co, Ni, Zn, Mn, Fe, etc.) with stoichiometric or even non-stoichiometric compositions, typically in a spinel structure, has recently attracted increasing research interest worldwide. Benefiting from their remarkable electrochemical properties, these MTMOs will play significant roles for low-cost and environmentally friendly energy storage/conversion technologies. In this Review, we summarize recent research advances in the rational design and efficient synthesis of MTMOs with controlled shapes, sizes, compositions, and micro-/nanostructures, along with their applications as electrode materials for lithium-ion batteries and electrochemical capacitors, and efficient electrocatalysts for the oxygen reduction reaction in metal-air batteries and fuel cells. Some future trends and prospects to further develop advanced MTMOs for next-generation electrochemical energy storage/conversion systems are also presented.

Anion-exchange membranes in electrochemical energy systems

Abstract: This article provides an up-to-date perspective on the use of anion-exchange membranes in fuel cells, electrolysers, redox flow batteries, reverse electrodialysis cells, and bioelectrochemical systems (e.g. microbial fuel cells). The aim is to highlight key concepts, misconceptions, the current state-of-the-art, technological and scientific limitations, and the future challenges (research priorities) related to the use of anion-exchange membranes in these energy technologies. All the references that the authors deemed relevant, and were available on the web by the manuscript submission date (30th April 2014), are included.

High lithium anodic performance of highly nitrogen-doped porous carbon prepared from a metal-organic framework

Abstract: Nitrogen-doped graphene can be used for lithium storage, but the nitrogen content is often quite low. Here, the authors synthesize graphene analogous with high nitrogen content using a zeolitic imidazolate framework, which show exceptional battery performances.

The tenth data release of the Sloan digital sky survey: First spectroscopic data from the SDSS-iii apache point observatory galactic evolution experiment

Abstract: The Sloan Digital Sky Survey (SDSS) has been in operation since 2000 April. This paper presents the Tenth Public Data Release (DR10) from its current incarnation, SDSS-III. This data release includes the first spectroscopic data from the Apache Point Observatory Galaxy Evolution Experiment (APOGEE), along with spectroscopic data from the Baryon Oscillation Spectroscopic Survey (BOSS) taken through 2012 July. The APOGEE instrument is a near-infrared R ~ 22,500 300 fiber spectrograph covering 1.514-1.696 μm. The APOGEE survey is studying the chemical abundances and radial velocities of roughly 100,000 red giant star candidates in the bulge, bar, disk, and halo of the Milky Way. DR10 includes 178,397 spectra of 57,454 stars, each typically observed three or more times, from APOGEE. Derived quantities from these spectra (radial velocities, effective temperatures, surface gravities, and metallicities) are also included. DR10 also roughly doubles the number of BOSS spectra over those included in the Ninth Data Release. DR10 includes a total of 1,507,954 BOSS spectra comprising 927,844 galaxy spectra, 182,009 quasar spectra, and 159,327 stellar spectra selected over 6373.2 deg2.

Oxygen vacancies confined in ultrathin indium oxide porous sheets for promoted visible-light water splitting.

Abstract: Finding an ideal model for disclosing the role of oxygen vacancies in photocatalysis remains a huge challenge. Herein, O-vacancies confined in atomically thin sheets is proposed as an excellent platform to study the O-vacancy–photocatalysis relationship. As an example, O-vacancy-rich/-poor 5-atom-thick In2O3 porous sheets are first synthesized via a mesoscopic-assembly fast-heating strategy, taking advantage of an artificial hexagonal mesostructured In-oleate complex. Theoretical/experimental results reveal that the O-vacancies endow 5-atom-thick In2O3 sheets with a new donor level and increased states of density, hence narrowing the band gap from the UV to visible regime and improving the carrier separation efficiency. As expected, the O-vacancy-rich ultrathin In2O3 porous sheets-based photoelectrode exhibits a visible-light photocurrent of 1.73 mA/cm2, over 2.5 and 15 times larger than that of the O-vacancy-poor ultrathin In2O3 porous sheets- and bulk In2O3-based photoelectrodes.

Face Alignment at 3000 FPS via Regressing Local Binary Features

Abstract: This paper presents a highly efficient, very accurate regression approach for face alignment. Our approach has two novel components: a set of local binary features, and a locality principle for learning those features. The locality principle guides us to learn a set of highly discriminative local binary features for each facial landmark independently. The obtained local binary features are used to jointly learn a linear regression for the final output. Our approach achieves the state-of-the-art results when tested on the current most challenging benchmarks. Furthermore, because extracting and regressing local binary features is computationally very cheap, our system is much faster than previous methods. It achieves over 3, 000 fps on a desktop or 300 fps on a mobile phone for locating a few dozens of landmarks.

Two dimensional nanomaterials for flexible supercapacitors.

Abstract: Flexible supercapacitors, as one of most promising emerging energy storage devices, are of great interest owing to their high power density with great mechanical compliance, making them very suitable as power back-ups for future stretchable electronics. Two-dimensional (2D) nanomaterials, including the quasi-2D graphene and inorganic graphene-like materials (IGMs), have been greatly explored to providing huge potential for the development of flexible supercapacitors with higher electrochemical performance. This review article is devoted to recent progresses in engineering 2D nanomaterials for flexible supercapacitors, which survey the evolution of electrode materials, recent developments in 2D nanomaterials and their hybrid nanostructures with regulated electrical properties, and the new planar configurations of flexible supercapacitors. Furthermore, a brief discussion on future directions, challenges and opportunities in this fascinating area is also provided.

Monolayer behaviour in bulk ReS2 due to electronic and vibrational decoupling

Abstract: Monolayers of transition metal dichalcogenides have emerged as interesting two-dimensional materials. Here, the authors show that in a new member of this family of compounds, rhenium disulphide, the layers in the bulk are vibrationally and electronically decoupled, so that they behave almost as monolayers.

Noble-Metal-Free Fe–N/C Catalyst for Highly Efficient Oxygen Reduction Reaction under Both Alkaline and Acidic Conditions

Abstract: In this work, we report the synthesis and assessment of a new non-precious-metal oxygen reduction reaction (ORR) catalyst from pyrolysis of an iron-coordinated complex which manifests superior activity in both alkaline and acidic media. 11,11′-bis(dipyrido[3,2-a:2′,3′-c]phenazinyl) (bidppz) was selected as a ligand for the formation of a nitrogen-rich iron-coordinated coordination polymer (Fe–bidppz) which forms a self-supporting catalyst containing high densities of nitrogen and iron doping by pyrolysis. The catalyst pyrolyzed at 800 °C (Fe–N/C-800) shows the highest ORR activity with onset and half-wave potentials of 923 and 809 mV in 0.1 M KOH, respectively, which are comparable to those of Pt/C (half-wave potential 818 mV vs RHE) at the same catalyst loading. Besides, the Fe–N/C-800 catalyst has an excellent ORR activity with onset and half-wave potentials only 38 and 59 mV less than those of the Pt/C catalyst in 0.1 M HClO4. The optimal Fe–N/C-800 catalyst displays much greater durability and toleran...

Low Overpotential in Vacancy-Rich Ultrathin CoSe2 Nanosheets for Water Oxidation

Abstract: According to Yang Shao-Horn's principle, CoSe2 is a promising candidate as an efficient, affordable, and sustainable alternative electrocatalyst for the oxygen evolution reaction, owing to its well-suited electronic configuration of Co ions. However, the catalytic efficiency of pure CoSe2 is still far below what is expected, because of its poor active site exposure yield. Herein, we successfully overcome the disadvantage of insufficient active sites in bulk CoSe2 by reducing its thickness into the atomic scale rather than any additional modification (such as doping or hybridizing with graphene or noble metals). The positron annihilation spectrometry and XAFS spectra provide clear evidence that a large number of VCo″ vacancies formed in the ultrathin nanosheets. The first-principles calculations reveal that these VCo″ vacancies can serve as active sites to efficiently catalyze the oxygen evolution reaction, manifesting an OER overpotential as low as 0.32 V at 10 mA cm(-2) in pH 13 medium, which is superior to the values for its bulk counterparts as well as those for the most reported Co-based electrocatalysts. Considering the outstanding performance of the simple, unmodified ultrathin CoSe2 nanosheets as the only catalyst, further improvement of the catalytic activity is expected when various strategies of doping or hybridizing are used. These results not only demonstrate the potential of a notable, affordable, and earth-abundant water oxidation electrocatalyst based on ultrathin CoSe2 nanosheets but also open up a promising avenue into the exploration of excellent active and durable catalysts toward replacing noble metals for oxygen electrocatalysis.

Titanium dioxide crystals with tailored facets

Abstract: Titanium dioxide (TiO2) has been the most intensively investigated binary transition metal oxide in the past four decades as indicated by Figure S1. Furthermore, the annual number of papers published on TiO2 has seen a continuous increase, particularly since the beginning of this century (Figure S2). This is understandable when one considers the wide range of applications of TiO2 from the conventional areas (i.e., pigment, cosmetic, toothpaste, and paint) to the later developed functional areas such as photoelectrochemical cell,(1-3) dye-sensitized solar cells (DSSCs),(4-11) photocatalysis,(12-24) catalysis,(25-31) photovoltaic cell,(32-34) lithium ion batteries,(35-41) sensors,(42-46) electron field emission,(47-51) microwave absorbing material, biomimetic growth, and biomedical treatments.(52-57) Nearly all these functional applications of TiO2 fall in the scope of energy, environment, and health, which are definitely the three most important and challenging themes facing the Human race that need to be addressed in this century. Besides the apparent merits including nontoxicity, elemental abundance, good chemical stability, and easy synthesis, TiO2 has attracted strong research interest worldwide due to its physicochemical properties for realizing various functions.(15, 58, 59) Especially, very encouraging progresses in photocatalysis and DSSCs with the involvement of TiO2 have greatly stimulated the rapid development of TiO2 crystals with controllable phase, size, shape, defect, and heteroatom.(58, 60-68)

An Experimental Study on Speech Enhancement Based on Deep Neural Networks

Abstract: This letter presents a regression-based speech enhancement framework using deep neural networks (DNNs) with a multiple-layer deep architecture. In the DNN learning process, a large training set ensures a powerful modeling capability to estimate the complicated nonlinear mapping from observed noisy speech to desired clean signals. Acoustic context was found to improve the continuity of speech to be separated from the background noises successfully without the annoying musical artifact commonly observed in conventional speech enhancement algorithms. A series of pilot experiments were conducted under multi-condition training with more than 100 hours of simulated speech data, resulting in a good generalization capability even in mismatched testing conditions. When compared with the logarithmic minimum mean square error approach, the proposed DNN-based algorithm tends to achieve significant improvements in terms of various objective quality measures. Furthermore, in a subjective preference evaluation with 10 listeners, 76.35% of the subjects were found to prefer DNN-based enhanced speech to that obtained with other conventional technique.

Single-layered ultrasmall nanoplates of MoS2 embedded in carbon nanofibers with excellent electrochemical performance for lithium and sodium storage.

Abstract: The preparation and electrochemical storage behavior of MoS2 nanodots--more precisely single-layered ultrasmall nanoplates--embedded in carbon nanowires has been studied. The preparation is achieved by an electrospinning process that can be easily scaled up. The rate performance and cycling stability of both lithium and sodium storage were found to be outstanding. The storage behavior is, moreover, highly exciting from a fundamental point of view, as the differences between the usual storage modes--insertion, conversion, interfacial storage--are beneficially blurred. The restriction to ultrasmall reaction domains allows for an almost diffusion-less and nucleation-free "conversion", thereby resulting in a high capacity and a remarkable cycling performance.

Breast cancer stem cells transition between epithelial and mesenchymal states reflective of their normal counterparts

Abstract: Previous studies have suggested that breast cancer stem cells (BCSCs) mediate metastasis, are resistant to radiation and chemotherapy, and contribute to relapse. Although several BCSC markers have been described, it is unclear whether these markers identify the same or independent BCSCs. Here, we show that BCSCs exist in distinct mesenchymal-like (epithelial-mesenchymal transition [EMT]) and epithelial-like (mesenchymal-epithelial transition [MET]) states. Mesenchymal-like BCSCs characterized as CD24−CD44+ are primarily quiescent and localized at the tumor invasive front, whereas epithelial-like BCSCs express aldehyde dehydrogenase (ALDH), are proliferative, and are located more centrally. The gene-expression profiles of mesenchymal-like and epithelial-like BCSCs are remarkably similar across different molecular subtypes of breast cancer, and resemble those of distinct basal and luminal stem cells found in the normal breast. We propose that the plasticity of BCSCs that allows them to transition between EMT- and MET-like states endows these cells with the capacity for tissue invasion, dissemination, and growth at metastatic sites.

On the Linear Convergence of the ADMM in Decentralized Consensus Optimization

Abstract: In decentralized consensus optimization, a connected network of agents collaboratively minimize the sum of their local objective functions over a common decision variable, where their information exchange is restricted between the neighbors. To this end, one can first obtain a problem reformulation and then apply the alternating direction method of multipliers (ADMM). The method applies iterative computation at the individual agents and information exchange between the neighbors. This approach has been observed to converge quickly and deemed powerful. This paper establishes its linear convergence rate for the decentralized consensus optimization problem with strongly convex local objective functions. The theoretical convergence rate is explicitly given in terms of the network topology, the properties of local objective functions, and the algorithm parameter. This result is not only a performance guarantee but also a guideline toward accelerating the ADMM convergence.

High-Performance Flexible Broadband Photodetector Based on Organolead Halide Perovskite

Abstract: Organolead halide perovskites have attracted extensive attentions as light harvesting materials for solar cells recently, because of its high charge-carrier mobilities, high photoconversion efficiencies, low energy cost, ease of deposition, and so on. Herein, with CH3NH3PbI3 film deposited on flexible ITO coated substrate, the first organolead halide perovskite based broadband photodetector is demonstrated. The organolead halide perovskite photodetector is sensitive to a broadband wavelength from the ultraviolet light to entire visible light, showing a photo-responsivity of 3.49 A W−1, 0.0367 A W−1, an external quantum efficiency of 1.19×103%, 5.84% at 365 nm and 780 nm with a voltage bias of 3 V, respectively. Additionally, the as-fabricated photodetector exhibit excellent flexibility and robustness with no obvious variation of photocurrent after bending for several times. The organolead halide perovskite photodetector with high sensitivity, high speed and broad spectrum photoresponse is promising for further practical applications. And this platform creates new opportunities for the development of low-cost, solution-processed and high-efficiency photodetectors.

Strongly coupled magnons and cavity microwave photons.

Abstract: We realize a cavity magnon-microwave photon system in which a magnetic dipole interaction mediates strong coupling between the collective motion of a large number of spins in a ferrimagnet and the microwave field in a three-dimensional cavity. By scaling down the cavity size and increasing the number of spins, an ultrastrong coupling regime is achieved with a cooperativity reaching 12 600. Interesting dynamic features including classical Rabi-like oscillation, magnetically induced transparency, and the Purcell effect are demonstrated in this highly versatile platform, highlighting its great potential for coherent information processing.

Generation of three-dimensional retinal tissue with functional photoreceptors from human iPSCs

Abstract: Induced pluripotent stem cells (iPSCs) hold great potential for modelling human developmental processes and diseases. Here the authors induce human iPSCs to spontaneously form fully laminated three-dimensional retinal tissue containing functional photoreceptor cells.

Tuning Interlayer Coupling in Large-Area Heterostructures with CVD-Grown MoS2 and WS2 Monolayers

Abstract: Band offsets between different monolayer transition metal dichalcogenides are expected to efficiently separate charge carriers or rectify charge flow, offering a mechanism for designing atomically thin devices and probing exotic two-dimensional physics. However, developing such large-area heterostructures has been hampered by challenges in synthesis of monolayers and effectively coupling neighboring layers. Here, we demonstrate large-area (>tens of micrometers) heterostructures of CVD-grown WS2 and MoS2 monolayers, where the interlayer interaction is externally tuned from noncoupling to strong coupling. Following this trend, the luminescence spectrum of the heterostructures evolves from an additive line profile where each layer contributes independently to a new profile that is dictated by charge transfer and band normalization between the WS2 and MoS2 layers. These results and findings open up venues to creating new material systems with rich functionalities and novel physical effects.

Doped graphene for metal-free catalysis.

Abstract: Graphene has attracted increasing attention in different scientific fields including catalysis. Via modification with foreign metal-free elements such as nitrogen, its unique electronic and spin structure can be changed and these doped graphene sheets have been successfully employed in some catalytic reactions recently, showing them to be promising catalysts for a wide range of reactions. In this review, we summarize the recent advancements of these new and interesting catalysts, with an emphasis on the universal origin of their catalytic mechanisms. We are full of hope for future developments, such as more precisely controlled doping methods, atom-scale surface characterization technology, generating more active catalysts via doping, and finding wide applications in many different fields.

Proton transport through one-atom-thick crystals

Abstract: Measurements show that monolayers of graphene and hexagonal boron nitride are unexpectedly highly permeable to thermal protons and that their conductivity rapidly increases with temperature, but that no proton transport is detected for few-layer crystals. A perfect graphene sheet is impermeable to all atoms and molecules: even hydrogen, the smallest of atoms, is not expected to penetrate through graphene's dense electronic cloud within billions of years. This characteristic is thought to extend to other two-dimensional crystals such as hexagonal boron nitride and molybdenum disulphide. Sheng Hu and colleagues now show that, surprisingly, monolayers of graphene and hexagonal boron nitride (but not molybdenum disulphide) are highly permeable to protons. In combination with their stability, this establishes these monolayers as promising candidates for use in many hydrogen-based technologies. Graphene is increasingly explored as a possible platform for developing novel separation technologies1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19. This interest has arisen because it is a maximally thin membrane that, once perforated with atomic accuracy, may allow ultrafast and highly selective sieving of gases, liquids, dissolved ions and other species of interest2,9,10,11,12,13,14,15,16,17,18,19. However, a perfect graphene monolayer is impermeable to all atoms and molecules under ambient conditions1,2,3,4,5,6,7: even hydrogen, the smallest of atoms, is expected to take billions of years to penetrate graphene’s dense electronic cloud3,4,5,6. Only accelerated atoms possess the kinetic energy required to do this20,21. The same behaviour might reasonably be expected in the case of other atomically thin crystals22,23. Here we report transport and mass spectroscopy measurements which establish that monolayers of graphene and hexagonal boron nitride (hBN) are highly permeable to thermal protons under ambient conditions, whereas no proton transport is detected for thicker crystals such as monolayer molybdenum disulphide, bilayer graphene or multilayer hBN. Protons present an intermediate case between electrons (which can tunnel easily through atomically thin barriers24) and atoms, yet our measured transport rates are unexpectedly high4,5 and raise fundamental questions about the details of the transport process. We see the highest room-temperature proton conductivity with monolayer hBN, for which we measure a resistivity to proton flow of about 10 Ω cm2 and a low activation energy of about 0.3 electronvolts. At higher temperatures, hBN is outperformed by graphene, the resistivity of which is estimated to fall below 10−3 Ω cm2 above 250 degrees Celsius. Proton transport can be further enhanced by decorating the graphene and hBN membranes with catalytic metal nanoparticles. The high, selective proton conductivity and stability make one-atom-thick crystals promising candidates for use in many hydrogen-based technologies.

Increasing the visible light absorption of graphitic carbon nitride (melon) photocatalysts by homogeneous self-modification with nitrogen vacancies.

Abstract: A novel reduced melon photocatalyst with a bandgap of 2.03 eV developed here has a widened visible light absorption range and suppressed radiative recombination of photo-excited charge carriers due to the homogeneous self-modification with nitrogen vacancies. As a consequence, the reduced melon shows a much superior photocatalytic activity compared to the pristine melon in generating •OH radicals and degrading the organic pollutant Rhodamine B.

Learning deep representations for graph clustering

Abstract: Recently deep learning has been successfully adopted in many applications such as speech recognition and image classification. In this work, we explore the possibility of employing deep learning in graph clustering. We propose a simple method, which first learns a nonlinear embedding of the original graph by stacked autoencoder, and then runs k-means algorithm on the embedding to obtain clustering result. We show that this simple method has solid theoretical foundation, due to the similarity between autoencoder and spectral clustering in terms of what they actually optimize. Then, we demonstrate that the proposed method is more efficient and flexible than spectral clustering. First, the computational complexity of autoencoder is much lower than spectral clustering: the former can be linear to the number of nodes in a sparse graph while the latter is super quadratic due to eigenvalue decomposition. Second, when additional sparsity constraint is imposed, we can simply employ the sparse autoencoder developed in the literature of deep learning; however, it is nonstraightforward to implement a sparse spectral method. The experimental results on various graph datasets show that the proposed method significantly outperforms conventional spectral clustering, which clearly indicates the effectiveness of deep learning in graph clustering.

Capacitance of carbon-based electrical double-layer capacitors

Abstract: Experimental electrical double-layer capacitances of porous carbon electrodes fall below ideal values, thus limiting the practical energy densities of carbon-based electrical double-layer capacitors. Here we investigate the origin of this behaviour by measuring the electrical double-layer capacitance in one to five-layer graphene. We find that the capacitances are suppressed near neutrality, and are anomalously enhanced for thicknesses below a few layers. We attribute the first effect to quantum capacitance effects near the point of zero charge, and the second to correlations between electrons in the graphene sheet and ions in the electrolyte. The large capacitance values imply gravimetric energy storage densities in the single-layer graphene limit that are comparable to those of batteries. We anticipate that these results shed light on developing new theoretical models in understanding the electrical double-layer capacitance of carbon electrodes, and on opening up new strategies for improving the energy density of carbon-based capacitors.

GeoMF: joint geographical modeling and matrix factorization for point-of-interest recommendation

Abstract: Point-of-Interest (POI) recommendation has become an important means to help people discover attractive locations However, extreme sparsity of user-POI matrices creates a severe challenge To cope with this challenge, viewing mobility records on location-based social networks (LBSNs) as implicit feedback for POI recommendation, we first propose to exploit weighted matrix factorization for this task since it usually serves collaborative filtering with implicit feedback better Besides, researchers have recently discovered a spatial clustering phenomenon in human mobility behavior on the LBSNs, ie, individual visiting locations tend to cluster together, and also demonstrated its effectiveness in POI recommendation, thus we incorporate it into the factorization model Particularly, we augment users' and POIs' latent factors in the factorization model with activity area vectors of users and influence area vectors of POIs, respectively Based on such an augmented model, we not only capture the spatial clustering phenomenon in terms of two-dimensional kernel density estimation, but we also explain why the introduction of such a phenomenon into matrix factorization helps to deal with the challenge from matrix sparsity We then evaluate the proposed algorithm on a large-scale LBSN dataset The results indicate that weighted matrix factorization is superior to other forms of factorization models and that incorporating the spatial clustering phenomenon into matrix factorization improves recommendation performance

Quantum error correction in a solid-state hybrid spin register

Abstract: Error correction is important in classical and quantum computation Decoherence caused by the inevitable interaction of quantum bits with their environment leads to dephasing or even relaxation Correction of the concomitant errors is therefore a fundamental requirement for scalable quantum computation Although algorithms for error correction have been known for some time, experimental realizations are scarce Here we show quantum error correction in a heterogeneous, solid-state spin system We demonstrate that joint initialization, projective readout and fast local and non-local gate operations can all be achieved in diamond spin systems, even under ambient conditions High-fidelity initialization of a whole spin register (99 per cent) and single-shot readout of multiple individual nuclear spins are achieved by using the ancillary electron spin of a nitrogen-vacancy defect Implementation of a novel non-local gate generic to our electron-nuclear quantum register allows the preparation of entangled states of three nuclear spins, with fidelities exceeding 85 per cent With these techniques, we demonstrate three-qubit phase-flip error correction Using optimal control, all of the above operations achieve fidelities approaching those needed for fault-tolerant quantum operation, thus paving the way to large-scale quantum computation Besides their use with diamond spin systems, our techniques can be used to improve scaling of quantum networks relying on phosphorus in silicon, quantum dots, silicon carbide or rare-earth ions in solids

Improving PARSEC models for very low mass stars

Abstract: Many stellar models present difficulties in reproducing basic observational relations of very low mass stars (VLMS), including the mass--radius relation and the optical colour--magnitudes of cool dwarfs. Here, we improve PARSEC models on these points. We implement the T--tau relations from PHOENIX BT-Settl model atmospheres as the outer boundary conditions in the PARSEC code, finding that this change alone reduces the discrepancy in the mass--radius relation from 8 to 5 per cent. We compare the models with multi--band photometry of clusters Praesepe and M67, showing that the use of T--tau relations clearly improves the description of the optical colours and magnitudes. But anyway, using both Kurucz and PHOENIX model spectra, model colours are still systematically fainter and bluer than the observations. We then apply a shift to the above T--tau relations, increasing from 0 at T_eff = 4730 K to ~14% at T_eff = 3160 K, to reproduce the observed mass--radius radius relation of dwarf stars. Taking this experiment as a calibration of the T--tau relations, we can reproduce the optical and near infrared CMDs of low mass stars in the old metal--poor globular clusters NGC6397 and 47Tuc, and in the intermediate--age and young solar--metallicity open clusters M67 and Praesepe. Thus, we extend PARSEC models using this calibration, providing VLMS models more suitable for the lower main sequence stars over a wide range of metallicities and wavelengths. Both sets of models are available on PARSEC webpage.