arXiv subjects

Haifeng Liu

Publications and source records attributed to Haifeng Liu.

68 records · Page 4Linked to original sources

arXiv2018

Deep Diabetologist: Learning to Prescribe Hyperglycemia Medications with Hierarchical Recurrent Neural Networks

In healthcare, applying deep learning models to electronic health records (EHRs) has drawn considerable attention. EHR data consist of a sequence of medical visits, i.e. a multivariate time series of diagnosis, medications, physical examinations, lab tests, etc. This sequential nature makes EHR well matching the power of Recurrent Neural Network (RNN). In this paper, we propose "Deep Diabetologist" - using RNNs for EHR sequential data modelling, to provide the personalized hyperglycemia medication prediction for diabetic patients. Particularly, we develop a hierarchical RNN to capture the heterogeneous sequential information in the EHR data. Our experimental results demonstrate the improved performance, compared with a baseline classifier using logistic regression. Moreover, hierarchical RNN models outperform basic ones, providing deeper data insights for clinical decision support.

cs.LG↗

arXiv2018

Addressing the Item Cold-start Problem by Attribute-driven Active Learning

In recommender systems, cold-start issues are situations where no previous events, e.g. ratings, are known for certain users or items. In this paper, we focus on the item cold-start problem. Both content information (e.g. item attributes) and initial user ratings are valuable for seizing users' preferences on a new item. However, previous methods for the item cold-start problem either 1) incorporate content information into collaborative filtering to perform hybrid recommendation, or 2) actively select users to rate the new item without considering content information and then do collaborative filtering. In this paper, we propose a novel recommendation scheme for the item cold-start problem by leverage both active learning and items' attribute information. Specifically, we design useful user selection criteria based on items' attributes and users' rating history, and combine the criteria in an optimization framework for selecting users. By exploiting the feedback ratings, users' previous ratings and items' attributes, we then generate accurate rating predictions for the other unselected users. Experimental results on two real-world datasets show the superiority of our proposed method over traditional methods.

cs.IR↗

arXiv2018

Deep Rotation Equivariant Network

Recently, learning equivariant representations has attracted considerable research attention. Dieleman et al. introduce four operations which can be inserted into convolutional neural network to learn deep representations equivariant to rotation. However, feature maps should be copied and rotated four times in each layer in their approach, which causes much running time and memory overhead. In order to address this problem, we propose Deep Rotation Equivariant Network consisting of cycle layers, isotonic layers and decycle layers. Our proposed layers apply rotation transformation on filters rather than feature maps, achieving a speed up of more than 2 times with even less memory overhead. We evaluate DRENs on Rotated MNIST and CIFAR-10 datasets and demonstrate that it can improve the performance of state-of-the-art architectures.

cs.CV↗

arXiv2018

PixelLink: Detecting Scene Text via Instance Segmentation

Most state-of-the-art scene text detection algorithms are deep learning based methods that depend on bounding box regression and perform at least two kinds of predictions: text/non-text classification and location regression. Regression plays a key role in the acquisition of bounding boxes in these methods, but it is not indispensable because text/non-text prediction can also be considered as a kind of semantic segmentation that contains full location information in itself. However, text instances in scene images often lie very close to each other, making them very difficult to separate via semantic segmentation. Therefore, instance segmentation is needed to address this problem. In this paper, PixelLink, a novel scene text detection algorithm based on instance segmentation, is proposed. Text instances are first segmented out by linking pixels within the same instance together. Text bounding boxes are then extracted directly from the segmentation result without location regression. Experiments show that, compared with regression-based methods, PixelLink can achieve better or comparable performance on several benchmarks, while requiring many fewer training iterations and less training data.

cs.CV↗

arXiv2017

On the Diversity of Realistic Image Synthesis

Many image processing tasks can be formulated as translating images between two image domains, such as colorization, super resolution and conditional image synthesis. In most of these tasks, an input image may correspond to multiple outputs. However, current existing approaches only show very minor diversity of the outputs. In this paper, we present a novel approach to synthesize diverse realistic images corresponding to a semantic layout. We introduce a diversity loss objective, which maximizes the distance between synthesized image pairs and links the input noise to the semantic segments in the synthesized images. Thus, our approach can not only produce diverse images, but also allow users to manipulate the output images by adjusting the noise manually. Experimental results show that images synthesized by our approach are significantly more diverse than that of the current existing works and equipping our diversity loss does not degrade the reality of the base networks.

cs.CV↗

arXiv2017

BENCHIP: Benchmarking Intelligence Processors

The increasing attention on deep learning has tremendously spurred the design of intelligence processing hardware. The variety of emerging intelligence processors requires standard benchmarks for fair comparison and system optimization (in both software and hardware). However, existing benchmarks are unsuitable for benchmarking intelligence processors due to their non-diversity and nonrepresentativeness. Also, the lack of a standard benchmarking methodology further exacerbates this problem. In this paper, we propose BENCHIP, a benchmark suite and benchmarking methodology for intelligence processors. The benchmark suite in BENCHIP consists of two sets of benchmarks: microbenchmarks and macrobenchmarks. The microbenchmarks consist of single-layer networks. They are mainly designed for bottleneck analysis and system optimization. The macrobenchmarks contain state-of-the-art industrial networks, so as to offer a realistic comparison of different platforms. We also propose a standard benchmarking methodology built upon an industrial software stack and evaluation metrics that comprehensively reflect the various characteristics of the evaluated intelligence processors. BENCHIP is utilized for evaluating various hardware platforms, including CPUs, GPUs, and accelerators. BENCHIP will be open-sourced soon.

cs.PF↗

arXiv2014

Personalized recommendation against crowd's popular selection

The problem of personalized recommendation in an ocean of data attracts more and more attention recently. Most traditional researches ignore the popularity of the recommended object, which resulting in low personality and accuracy. In this Letter, we proposed a personalized recommendation method based on weighted object network, punishing the recommended object that is the crowd's popular selection, namely, Anti-popularity index(AP), which can give enhanced personality, accuracy and diversity in contrast to mainstream baselines with a low computational complexity.

cs.IR↗

arXiv2014

The different roles of Pu-oxide overlayers in the hydrogenation of Pu-metal: An ab initio molecular dynamics study based on vdW-DFT+U

Based on the van der Waals density functional theory (vdW-DFT)+U scheme, we carry out the ab initio molecular dynamics (AIMD) study of the interaction dynamics for H$_{2}$ impingement against the stoichiometric PuO$_{2}$(111), the reduced PuO$_{2}$(111), and the stoichiometric $α$-Pu$_{2}$O$_{3}$(111) surfaces. The hydrogen molecular physisorption states, which can not be captured by pure DFT+\textit{U} method, are obtained by employing the vdW-DFT+\textit{U} scheme. We show that except for the weak physisorption, PuO$_{2}$(111) surfaces are so difficult of access that almost all of the H$_{2}$ molecules will bounce back to the vacuum when their initial kinetic energies are not sufficient. Although the dissociative adsorption of H$_{2}$ on PuO$_{2}$(111) surfaces is found to be very exothermic, the collision-induced dissociation barriers of H$_{2}$ are calculated to be as high as $3.2$ eV and $2.0$ eV for stoichiometric and reduced PuO$_{2}$ surfaces, respectively. Unlike PuO$_{2}$, our AIMD study directly reveals that the hydrogen molecules can penetrate into $α$-Pu$_{2}$O$_{3}$(111) surface and diffuse easily due to the $25$\ native O vacancies located along the $\langle $111$\rangle $ diagonals of $α$-Pu$_{2}$O$_{3}$ matrix. By examining the temperature effect and the internal vibrational excitations of H$_{2}$, we provide a detailed insight into the interaction dynamics of H$_{2}$ in $α$-Pu$_{2}$O$_{3}$. The optimum pathways for hydrogen penetration and diffusion, the corresponding energy barriers ($1.0$ eV and $0.53$ eV, respectively) and rate constants are systematically calculated. Overall, our study fairly reveals the different interaction mechanisms between H$_{2}$ and Pu-oxide surfaces, which have strong implications to the interpretation of experimental observations.

cond-mat.mtrl-sci↗

arXiv2013

An Energy-Efficient Localization Strategy for Smartphones

In recent years, smartphones have become prevalent. Much attention is being paid to developing and making use of mobile applications that require position information. The Global Positioning System (GPS) is a very popular localization technique used by these applications because of its high accuracy. However, GPS incurs an unacceptable energy consumption, which severely limits the use of smartphones and reduces the battery lifetime. Then an urgent requirement for these applications is a localization strategy that not only provides enough accurate position information to meet users' need but also consumes less energy. In this paper, we present an energy-efficient localization strategy for smartphone applications. On one hand, it can dynamically estimate the next localization time point to avoid unnecessary localization operations. On the other hand, it can also automatically select the energy-optimal localization method. We evaluate the strategy through a series of simulations. Experimental results show that it can significantly reduce the localization energy consumption of smartphones while ensuring a good satisfaction degree.

cs.NI↗

arXiv2012

First-principles study of surface properties of PuO2: Effects of thickness and O-vacancy on surface stability and chemical activity

The (111), (110), and (001) surfaces properties of PuO2 are studied by using density-functional theory+U method. The total-energy static calculations determine the relative order of stability for low-index PuO2 surfaces, namely, O-terminated (111) > (110) > defective (001) > polar (001). The effect of thickness is shown to modestly modulate the surface stability and chemical activity of the (110) surface. The high work function of 6.19 eV indicates the chemical inertia of the most stable (111) surface, and the surface O-vacancy with concentration C_V=25% can efficiently lower the work function to 4.35 eV, which is a crucial indicator of the difference in the surface chemical activities between PuO2 and α-Pu2O3. For the polar (001) surface, 50% on-surface O-vacancy can effectively quench the dipole moment and stabilize the surface structure, where the residual surface oxygen atoms are arranged in a zigzag manner along the <100> direction. We also investigate the relative stability of PuO2 surfaces in an oxygen environment. Under oxygen-rich conditions, the stoichiometric O-terminated (111) is found to be the most stable surface. Whereas under O-reducing conditions, the on-surface O-vacancy of C_V = 1/9 is stable, and for high reducing conditions, the (111) surface with nearly one monolayer subsurface oxygen removed (C_V = 8/9) becomes most stable.

cond-mat.mtrl-sci↗

arXiv2012

The environmental dependence of redox energetics of PuO2 and α-Pu2O3: A quantitative solution from DFT+U calculations

We report a comprehensive density functional theory (DFT) + $U$ study of the energetics of charged and neutral oxygen defects in both PuO$_{2}$ and $α$-Pu$_{2}$O$_{3}$, and present a quantitative determination of the equilibrium compositions of reduced PuO$_{2}$ (PuO$_{2-x}$) as functions of environmental temperature and partial pressure of oxygen, which shows fairly agreement with corresponding high-temperature experiments. Under ambient conditions, the endothermic reduction of PuO$_{2}$ to $α$-Pu$_{2}$O$_{3} $ is found to be facilitated by accompanying volume expansion of PuO$_{2-x}$ and the possible migration of O-vacancy, whereas further reduction of $α$-Pu$_{2}$O$_{3}$ is predicted to be much more difficult. In contrast to the endothermic oxidation of PuO$_{2}$,\ the oxidation of $α$-Pu$_{2} $O$_{3}$ is a stable exothermic process.

cond-mat.mtrl-sci↗

arXiv2008

Phase Transitions on Fixed Connected Graphs and Random Graphs in the Presence of Noise

In this paper, we study the phase transition behavior emerging from the interactions among multiple agents in the presence of noise. We propose a simple discrete-time model in which a group of non-mobile agents form either a fixed connected graph or a random graph process, and each agent, taking bipolar value either +1 or -1, updates its value according to its previous value and the noisy measurements of the values of the agents connected to it. We present proofs for the occurrence of the following phase transition behavior: At a noise level higher than some threshold, the system generates symmetric behavior (vapor or melt of magnetization) or disagreement; whereas at a noise level lower than the threshold, the system exhibits spontaneous symmetry breaking (solid or magnetization) or consensus. The threshold is found analytically. The phase transition occurs for any dimension. Finally, we demonstrate the phase transition behavior and all analytic results using simulations. This result may be found useful in the study of the collective behavior of complex systems under communication constraints.

math.OC↗

arXiv2007

Equation of state for shock-compressed porous molybdenum from first-principles mean-field potential calculations

The Hugoniot curves for shock-compressed molybdenum with initial porosities of 1.0, 1.26, 1.83, and 2.31 are theoretically investigated. The method of calculations combines the first-principles treatment for zero- and finite-temperature electronic contribution and the mean-field-potential approach for the ion-thermal contribution to the total free energy. Our calculated results reproduce the Hugoniot properties of porous molybdenum quite well. At low porosity, in particular, the calculations show a complete agreement with the experimental measurements over the full range of data. For the two large porosity values of 1.83 and 2.31, our results are well in accord with the experimental data points up to the particle velocity of 3.5 km/s, and tend to overestimate the shock-wave velocity and Hugoniot pressure when further increasing the particle velocity. In addition, the temperature along the principal Hugoniot is also extensively investigated for porous molybdenum.

cond-mat.mtrl-sci↗

arXiv2007

Mean-field potential calculations of high-pressure equation of state for shock-compressed BeO

A systematic study of the Hugoniot equation of state, phase transition, and the other thermodynamic properties including the Hugoniot temperature, the electronic and ionic heat capacities, and the Grüneisen parameter for shock-compressed BeO, is presented by calculating the total free energy. The method of calculations combines first-principles treatment for 0-K and finite-T electronic contribution and the mean-field-potential approach for the vibrational contribution of the lattice ion to the total energy. Our calculated Hugoniot shows good agreement with the experimental data.

cond-mat.mtrl-sci↗