arXiv subjects

Li Ma

Publications and source records attributed to Li Ma.

At least 145 records · Page 8Linked to original sources

arXiv2020

Efficient in-situ image and video compression through probabilistic image representation

Fast and effective image compression for multi-dimensional images has become increasingly important for efficient storage and transfer of massive amounts of high-resolution images and videos. Desirable properties in compression methods include (1) high reconstruction quality at a wide range of compression rates while preserving key local details, (2) computational scalability, (3) applicability to a variety of different image/video types and of different dimensions, (4) progressive transmission, and (5) ease of tuning. We present such a method for multi-dimensional image compression called Compression via Adaptive Recursive Partitioning (CARP). CARP uses an optimal permutation of the image pixels inferred from a Bayesian probabilistic model on recursive partitions of the image to reduce its effective dimensionality, achieving a parsimonious representation that preserves information. CARP uses a multi-layer Bayesian hierarchical model to achieve in-situ compression along with self-tuning and regularization, with just one single parameter to be specified by the user to achieve the desired compression rate. Extensive numerical experiments using a variety of datasets including 2D still images, real-life YouTube videos, and surveillance videos show that CARP dominates the state-of-the-art image/video compression approaches---including JPEG, JPEG2000, BPG, MPEG4, HEVC and a neural network-based method---for all of these different image types and on nearly all of the individual images and videos over some methods.

eess.IV↗

arXiv2020

Learning Asymmetric and Local Features in Multi-Dimensional Data through Wavelets with Recursive Partitioning

Effective learning of asymmetric and local features in images and other data observed on multi-dimensional grids is a challenging objective critical for a wide range of image processing applications involving biomedical and natural images. It requires methods that are sensitive to local details while fast enough to handle massive numbers of images of ever increasing sizes. We introduce a probabilistic model-based framework that achieves these objectives by incorporating adaptivity into discrete wavelet transforms (DWT) through Bayesian hierarchical modeling, thereby allowing wavelet bases to adapt to the geometric structure of the data while maintaining the high computational scalability of wavelet methods---linear in the sample size (e.g., the resolution of an image). We derive a recursive representation of the Bayesian posterior model which leads to an exact message passing algorithm to complete learning and inference. While our framework is applicable to a range of problems including multi-dimensional signal processing, compression, and structural learning, we illustrate its work and evaluate its performance in the context of image reconstruction using real images from the ImageNet database, two widely used benchmark datasets, and a dataset from retinal optical coherence tomography and compare its performance to state-of-the-art methods based on basis transforms and deep learning.

stat.ME↗

arXiv2020

Dirichlet-tree multinomial mixtures for clustering microbiome compositions

Studying the human microbiome has gained substantial interest in recent years, and a common task in the analysis of these data is to cluster microbiome compositions into subtypes. This subdivision of samples into subgroups serves as an intermediary step in achieving personalized diagnosis and treatment. In applying existing clustering methods to modern microbiome studies including the American Gut Project (AGP) data, we found that this seemingly standard task, however, is very challenging in the microbiome composition context due to several key features of such data. Standard distance-based clustering algorithms generally do not produce reliable results as they do not take into account the heterogeneity of the cross-sample variability among the bacterial taxa, while existing model-based approaches do not allow sufficient flexibility for the identification of complex within-cluster variation from cross-cluster variation. Direct applications of such methods generally lead to overly dispersed clusters in the AGP data and such phenomenon is common for other microbiome data. To overcome these challenges, we introduce Dirichlet-tree multinomial mixtures (DTMM) as a Bayesian generative model for clustering amplicon sequencing data in microbiome studies. DTMM models the microbiome population with a mixture of Dirichlet-tree kernels that utilizes the phylogenetic tree to offer a more flexible covariance structure in characterizing within-cluster variation, and it provides a means for identifying a subset of signature taxa that distinguish the clusters. We perform extensive simulation studies to evaluate the performance of DTMM and compare it to state-of-the-art model-based and distance-based clustering methods in the microbiome context. Finally, we report a case study on the fecal data from the AGP to identify compositional clusters among individuals with inflammatory bowel disease and diabetes.

stat.ME↗

arXiv2020

Coarsened mixtures of hierarchical skew normal kernels for flow cytometry analyses

Flow cytometry (FCM) is the standard multi-parameter assay for measuring single cell phenotype and functionality. It is commonly used for quantifying the relative frequencies of cell subsets in blood and disaggregated tissues. A typical analysis of FCM data involves cell classification---that is, the identification of cell subgroups in the sample---and comparisons of the cell subgroups across samples or conditions. While modern experiments often necessitate the collection and processing of samples in multiple batches, analysis of FCM data across batches is challenging because differences across samples may occur due to either true biological variation or technical reasons such as antibody lot effects or instrument optics across batches. Thus a critical step in comparative analyses of multi-sample FCM data---yet missing in existing automated methods for analyzing such data---is cross-sample calibration, whose goal is to align corresponding cell subsets across multiple samples in the presence of technical variations, so that biological variations can be meaningfully compared. We introduce a Bayesian nonparametric hierarchical modeling approach for accomplishing both calibration and cell classification simultaneously in a unified probabilistic manner. Three important features of our method make it particularly effective for analyzing multi-sample FCM data: a nonparametric mixture avoids prespecifying the number of cell clusters; a hierarchical skew normal kernel that allows flexibility in the shapes of the cell subsets and cross-sample variation in their locations; and finally the "coarsening" strategy makes inference robust to departures from the model such as heavy-tailness not captured by the skew normal kernels. We demonstrate the merits of our approach in simulated examples and carry out a case study in the analysis of two multi-sample FCM data sets.

stat.AP↗

arXiv2020

Constants and heat flow on graphs

In this article, we first introduce the concepts of vector fields and their divergence, and we recall the concepts of the gradient, Laplacian operator, Cheeger constants, eigenvalues, and heat kernels on a locally finite graph $V$. We give a projective characteristic of the eigenvalues. We also give an extension of Barta Theorem. Then we introduce the mini-max value of a function on a locally finite and locally connected graph. We show that for a coercive function on on a locally finite and locally connected graph, there is a mini-max value of the function provided it has two strict local minima values. We consider the discrete Morse flow for the heat flow on a finite graph in the locally finite graph $V$. We show that under suitable assumptions on the graph one has a weak discrete Morse flow for the heat flow on $S$ on any time interval. We also study the heat flow with time-variable potential and its discrete Morse flow. We propose the concepts of harmonic maps from a graph to a Riemannian manifold and pose some open questions.

math.AP↗

arXiv2020

Tritonlike molecules of three identical baryons

In nuclear physics, triton and helium-3 nucleus can be understood as three-body hadronic molecules. Analogous to the loosely bound structures for the triton and helium-3 nucleus, whether there is a bound state formed by three hadrons leaves us an open issue. Based on the one-boson exchange model as well as the adoption of the variational approach, we make a comprehensive investigation on the tritonlike systems of three identical baryons $NNN$, $ΛΛΛ$, $ΞΞΞ$ and $ΣΣΣ$. We predict that the three-body molecular states for the systems of three identical hadrons of baryon octet are probably existent as long as their two-body subsystems have bound states. The numerical results of this work may be helpful for the theoretical and experimental researches on the tri-hadron molecules in future.

hep-ph↗

arXiv2020

High-throughput GPU layered decoder of multi-edge type low density parity check codes in continuous-variable quantum key distribution systems

The decoding throughput in the postprocessing is one of the bottlenecks for a continuous-variable quantum key distribution (CV-QKD) system. In this paper, we propose a layered decoder to decode quasi-cyclic multi-edge type LDPC (QC-METLDPC) codes based on graphic processing unit (GPU) in continuous-variable quantum key distribution (CV-QKD) systems. We optimize the storage method of the parity check matrix, merge the sub-matrices which are unrelated, and decode multiple codewords in parallel on GPU. Simulation results demonstrate that the average decoding speed of LDPC codes with three typical code rates, i.e., 0.1, 0.05 and 0.02, is up to 64.11Mbits/s, 48.65Mbits/s and 39.51Mbits/s, respectively, when decoding 128 codewords of length 106 simultaneously without early termination.

quant-ph↗

arXiv2020

The gamma-ray and Optical Variability Analysis of the BL Lac Object 3FGL J0449.4-4350

We have assembled the historical light curves of the BL Lac Object 3FGL J0449.4-4350 at optical and gamma-ray bands, the time spanning about 10 years, analyzed the periodic variability of the light curves by using four different methods (Lomb-Scargle periodogram, REDFIT38, Jurkevich and DACF). We detected a marginally possible quasi-periodic oscillation (QPO) of ~450 days. Assuming it originates from the helical motion jet in a supermassive binary black hole (SMBBH) system undergoing major merger, we estimate the primary black hole mass M~7.7*10^{9} M_sun. To explore the origin of the gamma-ray, we investigated the optical-gamma-ray correlations using discrete correlation function (DCF) method, and found that the correlation between the two bands is very significant. This strong correlation tends to imply lepton self-synchro-Compton (LSSC) model to produce the gamma-ray.

astro-ph.HE↗

arXiv2020

Privacy-preserving collaborative machine learning on genomic data using TensorFlow

Machine learning (ML) methods have been widely used in genomic studies. However, genomic data are often held by different stakeholders (e.g. hospitals, universities, and healthcare companies) who consider the data as sensitive information, even though they desire to collaborate. To address this issue, recent works have proposed solutions using Secure Multi-party Computation (MPC), which train on the decentralized data in a way that the participants could learn nothing from each other beyond the final trained model. We design and implement several MPC-friendly ML primitives, including class weight adjustment and parallelizable approximation of activation function. In addition, we develop the solution as an extension to TF Encrypted~\citep{dahl2018private}, enabling us to quickly experiment with enhancements of both machine learning techniques and cryptographic protocols while leveraging the advantages of TensorFlow's optimizations. Our implementation compares favorably with state-of-the-art methods, winning first place in Track IV of the iDASH2019 secure genome analysis competition.

cs.CR↗

arXiv2020

Passivation mechanisms and pre-oxidation effects on model surfaces of FeCrNi austenitic stainless steel

Passivation mechanisms were investigated on (100)-oriented Fe-18Cr-13Ni surfaces with direct transfer between surface preparation and analysis by X-ray photoelectron spectroscopy and scanning tunneling microscopy and electrochemical characterization. Starting from oxide-free surfaces, pre-oxidation at saturation under ultra-low pressure (ULP) oxygen markedly promotes the oxide film Cr(III) enrichment and hinders/delays subsequent iron oxidation in water-containing environment. Exposure to sulfuric acid at open circuit potential causes preferential dissolution of oxidized iron species. Anodic passivation forces oxide film re-growth, Cr(III) dehydroxylation and further enrichment. ULP pre-oxidation promotes Cr(III) hydroxide formation at open circuit potential, compactness of the nanogranular oxide film and corrosion protection.

cond-mat.mtrl-sci↗

arXiv2019

Tri-meson bound state $BBB^{\ast}$ via delocalized $π$~Bond

During the last decades, numerous exotic states which cannot be explained by the conventional quark model have been observed in experiment. Some of them can be understood as two-body hadronic molecules, such as the famous $X(3872)$, analogous to deuteron in nuclear physics. Along the same line, the existence of the triton leaves an open question whether there is a bound state formed by three hadrons. Since, for a given potential, a system with large reduced masses is more easier to form a bound state, we study the $BBB^{\ast}$ system with the one-pion exchange potential as an exploratory step by solving the three-body Schrödinger Equation. We predict that a tri-meson molecular state for the $BBB^{\ast}$ system is probably existent as long as the molecular states of its two-body subsystem $BB^*$ exist.

hep-ph↗

arXiv2019

The strong decay patterns of $Z_c$ and $Z_b$ states in the relativized quark model

Employing the relativized quark model and the quark-interchange model, we investigate the decay of the charged heavy quarkonium-like states $Z_c(3900)$, $Z_c(4020)$, $Z_c(4430)$, $Z_b(10610)$ and $Z_b(10650)$ into the ground and radially excited heavy quarkonia via emitting a pion meson. The $Z_c$ and $Z_b$ states are assumed to be hadronic molecules composed of open-flavor heavy mesons. The calculated decay ratios can be compared with the experimental data, which are useful in judging whether the molecule state assignment for the corresponding $Z_c$ or $Z_b$ state is reasonable or not. The theoretical framework constructed in this work will be helpful in revealing the underlying structures of some exotic hadrons.

hep-ph↗

arXiv2019

G-flocking: Flocking Model Optimization based on Genetic Framework

Flocking model has been widely used to control robotic swarm. However, with the increasing scalability, there exist complex conflicts for robotic swarm in autonomous navigation, brought by internal pattern maintenance, external environment changes, and target area orientation, which results in poor stability and adaptability. Hence, optimizing the flocking model for robotic swarm in autonomous navigation is an important and meaningful research domain.

cs.MA↗

arXiv2019

Stainless Steel Surface Structure and Initial Oxidation at Nanometric and Atomic Scales

The durability of passivable metals and alloys is often limited by the stability of the surface oxide film, the passive film, providing self-protection against corrosion in aggressive environments. Improving this stability requires to develop a deeper understanding of the surface structure and initial surface reactivity at the nanometric or atomic scale. In this work we applied scanning tunneling microscopy to unravel the surface structure of a model stainless steel surface in the metallic state and its local modifications induced by initial reaction in dioxygen gas. The results show a rich and complex structure of the oxide-free surface with reconstructed atomic lattice and self-organized lines of surface vacancies at equilibrium. New insight is brought into the mechanisms of initial oxidation at steps and vacancy injection on terraces leading to Cr-rich oxide nuclei and locally Cr-depleted terraces, impacting the subsequent mechanism of chromium enrichment essential to the stability of the surface oxide.

cond-mat.mtrl-sci↗

arXiv2019

A Bayesian hierarchical model for related densities using Polya trees

Bayesian hierarchical models are used to share information between related samples and obtain more accurate estimates of sample-level parameters, common structure, and variation between samples. When the parameter of interest is the distribution or density of a continuous variable, a hierarchical model for continuous distributions is required. A number of such models have been described in the literature using extensions of the Dirichlet process and related processes, typically as a distribution on the parameters of a mixing kernel. We propose a new hierarchical model based on the Pólya tree, which allows direct modeling of densities and enjoys some computational advantages over the Dirichlet process. The Pólya tree also allows more flexible modeling of the variation between samples, providing more informed shrinkage and permitting posterior inference on the dispersion function, which quantifies the variation among sample densities. We also show how the model can be extended to cluster samples in situations where the observed samples are believed to have been drawn from several latent populations.

stat.ME↗

arXiv2019

Bayesian graphical compositional regression for microbiome data

An important task in microbiome studies is to test the existence of and give characterization to differences in the microbiome composition across groups of samples. Important challenges of this problem include the large within-group heterogeneities among samples and the existence of potential confounding variables that, when ignored, increase the chance of false discoveries and reduce the power for identifying true differences. We propose a probabilistic framework to overcome these issues by combining three ideas: (i) a phylogenetic tree-based decomposition of the cross-group comparison problem into a series of local tests, (ii) a graphical model that links the local tests to allow information sharing across taxa, and (iii) a Bayesian testing strategy that incorporates covariates and integrates out the within-group variation, avoiding potentially unstable point estimates. We derive an efficient inference algorithm based on numerical integration and junction-tree message passing, conduct extensive simulation studies to investigate the performance of our approach, and compare it to state-of-the-art methods in a number of representative settings. We then apply our method to the American Gut data to analyze the association of dietary habits and human's gut microbiome composition in the presence of covariates, and illustrate the importance of incorporating covariates in microbiome cross-group comparison.

stat.ME↗

arXiv2019

Absence of Metallic Behavior in Epitaxial NiCo2O4 Thin Films: Role of Microstructural Disorder

Despite the low resistivity (~ 1 mohm cm), the metallic electrical transport has not been commonly observed in the inverse spinel NiCo2O4, except in certain epitaxial thin films. Previous studies have stressed the effect of valence mixing and degree of spinel inversion on the electric conduction of NiCo2O4 films. In this work, we have studied the effect of microstructure by comparing the NiCo2O4 epitaxial films grown on MgAl2O4 (111) and on Al2O3 (0001) substrates. Although the optimal growth condition and the magnetic properties are similar for the NiCo2O4/MgAl2O4 and the NiCo2O4/Al2O3, they show metallic and semiconducting electrical transport respectively. Despite similar temperature and field dependence of magnetization, the NiCo2O4/Al2O3 show much larger magnetoresistance at low temperature. Post-growth annealing decreases the resistivity of NiCo2O4/Al2O3, but the annealed films are still semiconducting. The correlation between the structural correlation length and the resistivity suggests that the microstructural disorder, generated by the dramatic mismatch between the NiCo2O4 and Al2O3 crystal structures, may be the origin of the absence of the metallic electrical transport in NiCo2O4. These results reveal microstructural disorder as another key factor in controlling the electrical transport of NiCo2O4, with potentially large magnetoresistance for spintronics application.

cond-mat.mtrl-sci↗

arXiv2019

POI Semantic Model with a Deep Convolutional Structure

When using the electronic map, POI retrieval is the initial and important step, whose quality directly affects the user experience. Similarity between user query and POI information is the most critical feature in POI retrieval. An accurate similarity calculation is challenging since the mismatch between a query and a retrieval text may exist in the case of a mistyped query or an alias inquiry. In this paper, we propose a POI latent semantic model based on deep networks, which can effectively extract query features and POI information features for the similarity calculation. Our model describes the semantic information of complex texts at multiple layers, and achieves multi-field matches by modeling POI's name and detailed address respectively. Our model is evaluated by the POI retrieval ranking datasets, including the labeled data of relevance and real-world user click data in POI retrieval. Results show that our model significantly outperforms our competitors in POI retrieval ranking tasks. The proposed algorithm has become a critical component of an online system serving millions of people everyday.

cs.IR↗