Searcharxiv⌕ Search

arXiv subjects

Boyu Zhang

Publications and source records attributed to Boyu Zhang.

At least 73 records · Page 4Linked to original sources
arXiv2021

Predicting Material Properties Using a 3D Graph Neural Network with Invariant Local Descriptors

Accurate prediction of physical properties is critical for discovering and designing novel materials. Machine learning technologies have attracted significant attention in the materials science community for their potential for large-scale screening. Graph Convolution Neural Network (GCNN) is one of the most successful machine learning methods because of its flexibility and effectiveness in describing 3D structural data. Most existing GCNN models focus on the topological structure but overly simplify the three-dimensional geometric structure. However, in materials science, the 3D-spatial distribution of atoms is crucial for determining the atomic states and interatomic forces. This paper proposes an adaptive GCNN with a novel convolution mechanism that simultaneously models atomic interactions among all neighbor atoms in three-dimensional space. We apply the proposed model to two distinctly challenging problems on predicting material properties. The first is Henry's constant for gas adsorption in Metal-Organic Frameworks (MOFs), which is notoriously difficult because of its high sensitivity to atomic configurations. The second is the ion conductivity in solid-state crystal materials, which is difficult because of few labeled data available for training. The new model outperforms existing graph-based models on both data sets, suggesting that the critical three-dimensional geometric information is indeed captured.

cond-mat.mtrl-sci↗
arXiv2021

BUSIS: A Benchmark for Breast Ultrasound Image Segmentation

Breast ultrasound (BUS) image segmentation is challenging and critical for BUS Comput-er-Aided Diagnosis (CAD) systems. Many BUS segmentation approaches have been studied in the last two decades, but the performances of most approaches have been assessed using relatively small private datasets with different quantitative metrics, which results in a discrepancy in performance comparison. Therefore, there is a pressing need for building a benchmark to compare existing methods using a public dataset objectively, to determine the performance of the best breast tumor segmentation algorithm available today, and to investigate what segmentation strategies are valuable in clinical practice and theoretical study. In this work, a benchmark for B-mode breast ultrasound image segmentation is presented. In the benchmark, 1) we collected 562 breast ultrasound images, prepared a software tool, and involved four radiologists in obtaining accurate annotations through standardized procedures; 2) we extensively compared the performance of sixteen state-of-the-art segmentation methods and discussed their advantages and disadvantages; 3) we proposed a set of valuable quantitative metrics to evaluate both semi-automatic and fully automatic segmentation approaches; and 4) the successful segmentation strategies and possible future improvements are discussed in details.

cs.CV↗
arXiv2021

BI-RADS-Net: An Explainable Multitask Learning Approach for Cancer Diagnosis in Breast Ultrasound Images

In healthcare, it is essential to explain the decision-making process of machine learning models to establish the trustworthiness of clinicians. This paper introduces BI-RADS-Net, a novel explainable deep learning approach for cancer detection in breast ultrasound images. The proposed approach incorporates tasks for explaining and classifying breast tumors, by learning feature representations relevant to clinical diagnosis. Explanations of the predictions (benign or malignant) are provided in terms of morphological features that are used by clinicians for diagnosis and reporting in medical practice. The employed features include the BI-RADS descriptors of shape, orientation, margin, echo pattern, and posterior features. Additionally, our approach predicts the likelihood of malignancy of the findings, which relates to the BI-RADS assessment category reported by clinicians. Experimental validation on a dataset consisting of 1,192 images indicates improved model accuracy, supported by explanations in clinical terms using the BI-RADS lexicon.

cs.CV↗
arXiv2021

On Floer minimal knots in sutured manifolds

Suppose $(M, γ)$ is a balanced sutured manifold and $K$ is a rationally null-homologous knot in $M$. It is known that the rank of the sutured Floer homology of $M\backslash N(K)$ is at least twice the rank of the sutured Floer homology of $M$. This paper studies the properties of $K$ when the equality is achieved for instanton homology. As an application, we show that if $L\subset S^3$ is a fixed link and $K$ is a knot in the complement of $L$, then the instanton link Floer homology of $L\cup K$ achieves the minimum rank if and only if $K$ is the unknot in $S^3\backslash L$.

math.GT↗
arXiv2021

Statistical Mechanical Analysis of Neural Network Pruning

Deep learning architectures with a huge number of parameters are often compressed using pruning techniques to ensure computational efficiency of inference during deployment. Despite multitude of empirical advances, there is a lack of theoretical understanding of the effectiveness of different pruning methods. We inspect different pruning techniques under the statistical mechanics formulation of a teacher-student framework and derive their generalization error (GE) bounds. It has been shown that Determinantal Point Process (DPP) based node pruning method is notably superior to competing approaches when tested on real datasets. Using GE bounds in the aforementioned setup we provide theoretical guarantees for their empirical observations. Another consistent finding in literature is that sparse neural networks (edge pruned) generalize better than dense neural networks (node pruned) for a fixed number of parameters. We use our theoretical setup to prove this finding and show that even the baseline random edge pruning method performs better than the DPP node pruning method. We also validate this empirically on real datasets.

cs.LG↗
arXiv2021

Monopoles and foliations without holonomy-invariant transverse measure

This article proves a uniform exponential decay estimate for Seiberg-Witten equations on non-compact 4-manifolds with exact symplectic ends of bounded geometry. This is an extension of the analysis for asymptotically flat almost Kähler (AFAK) structures by Kronheimer and Mrowka. As an application, we construct an invariant for smooth foliations without holonomy-invariant transverse measure, which takes value in the boundary-stable version of the monopole Floer homology group, without invoking the Eliashberg-Thurston perturbation.

math.GT↗
arXiv2021

On meridian-traceless SU(2)-representations of link groups

Suppose L is a link in $S^3$. We show that $π_1(S^3-L)$ admits an irreducible meridian-traceless representation in SU(2) if and only if L is not the unknot, the Hopf link, or a connected sum of Hopf links. As a corollary, $π_1(S^3-L)$ admits an irreducible representation in SU(2) if and only if L is neither the unknot nor the Hopf link. This result generalizes a theorem of Kronheimer and Mrowka to the case of links.

math.GT↗
arXiv2021

On the compactness problem for a family of generalized Seiberg-Witten equations in dimension three

We prove an abstract compactness theorem for a family of generalized Seiberg-Witten equations in dimension three. This result recovers Taubes' compactness theorem for stable flat $\mathbf{P}\mathrm{SL}_2(\mathbf{C})$-connections as well as the compactness theorem for Seiberg-Witten equations with multiple spinors. Furthermore, this result implies a compactness theorem for the ADHM$_{1,2}$ Seiberg-Witten equation, which partially verifies a conjecture by Doan and Walpuski.

math.DG↗
arXiv2021

Classification of links with Khovanov homology of minimal rank

If L is an oriented link with $n$ components, then the rank of its Khovanov homology is at least $2^n$. We classify all the links whose Khovanov homology with Z/2-coefficients achieves this lower bound, and show that such links can be obtained by iterated connected sums and disjoint unions of Hopf links and unknots. This gives a positive answer to a question asked by Batson and Seed.

math.GT↗
arXiv2020

All-optical helicity-independent switching state diagram in GdFeCo alloys

Ultra-fast magnetization switching induced by a single femtosecond laser pulse, under no applied magnetic field has attracted a lot of attention during the last 10 years because of its high potential for low energy and ultra-fast memory applications. Single-pulse helicity-independent switching has mostly been demonstrated for Gd based materials. It is now important to optimize the pulse duration and the energy needed to switch a GdFeCo magnet depending on the alloy thickness, concentration. Here we experimentally report state diagrams showing the magnetic state obtained after one single pulse depending on the laser pulse duration and fluence for various GdFeCo thin films with different compositions and thicknesses. We demonstrate that these state diagrams share similar characteristics: the fluence window for switching narrows for longer pulse duration and for the considered pulse duration range the critical fluence for single pulse switching increases linearly as a function of the pulse duration while the critical fluence required for creating a multidomain state remains almost constant. Calculations based on the atomistic spin model qualitatively reproduce the experimental state diagrams and their evolution. By studying the effect of the composition and the thickness on the state diagram, we demonstrated that the best energy efficiency and the longest pulse duration for switching are obtained for concentration around the magnetic compensation.

physics.optics↗
arXiv2020

Individual-level Anxiety Detection and Prediction from Longitudinal YouTube and Google Search Engagement Logs

Anxiety disorder is one of the world's most prevalent mental health conditions, arising from complex interactions of biological and environmental factors and severely interfering one's ability to lead normal life activities. Current methods for detecting anxiety heavily rely on in-person interviews, which can be expensive, time-consuming, and blocked by social stigmas. In this work, we propose an alternative method to identify individuals with anxiety and further estimate their levels of anxiety using personal online activity histories from YouTube and the Google Search engine, platforms that are used by millions of people daily. We ran a longitudinal study and collected multiple rounds of anonymized YouTube and Google Search logs from volunteering participants, along with their clinically validated ground-truth anxiety assessment scores. We then developed explainable features that capture both the temporal and contextual aspects of online behaviors. Using those, we were able to train models that (i) identify individuals having anxiety disorder with an average F1 score of 0.83 and (ii) assess the level of anxiety by predicting the gold standard Generalized Anxiety Disorder 7-item scores (ranges from 0 to 21) with a mean square error of 1.87 based on the ubiquitous individual-level online engagement data. Our proposed anxiety assessment framework is cost-effective, time-saving, scalable, and opens the door for it to be deployed in real-world clinical settings, empowering care providers and therapists to learn about anxiety disorders of patients non-invasively at any moment in time.

cs.HC↗
arXiv2020

A Use of Even Activation Functions in Neural Networks

Despite broad interest in applying deep learning techniques to scientific discovery, learning interpretable formulas that accurately describe scientific data is very challenging because of the vast landscape of possible functions and the "black box" nature of deep neural networks. The key to success is to effectively integrate existing knowledge or hypotheses about the underlying structure of the data into the architecture of deep learning models to guide machine learning. Currently, such integration is commonly done through customization of the loss functions. Here we propose an alternative approach to integrate existing knowledge or hypotheses of data structure by constructing custom activation functions that reflect this structure. Specifically, we study a common case when the multivariate target function $f$ to be learned from the data is partially exchangeable, \emph{i.e.} $f(u,v,w)=f(v,u,w)$ for $u,v\in \mathbb{R}^d$. For instance, these conditions are satisfied for the classification of images that is invariant under left-right flipping. Through theoretical proof and experimental verification, we show that using an even activation function in one of the fully connected layers improves neural network performance. In our experimental 9-dimensional regression problems, replacing one of the non-symmetric activation functions with the designated "Seagull" activation function $\log(1+x^2)$ results in substantial improvement in network performance. Surprisingly, even activation functions are seldom used in neural networks. Our results suggest that customized activation functions have great potential in neural networks.

cs.LG↗
arXiv2020

Detecting Individuals with Depressive Disorder fromPersonal Google Search and YouTube History Logs

Depressive disorder is one of the most prevalent mental illnesses among the global population. However, traditional screening methods require exacting in-person interviews and may fail to provide immediate interventions. In this work, we leverage ubiquitous personal longitudinal Google Search and YouTube engagement logs to detect individuals with depressive disorder. We collected Google Search and YouTube history data and clinical depression evaluation results from $212$ participants ($99$ of them suffered from moderate to severe depressions). We then propose a personalized framework for classifying individuals with and without depression symptoms based on mutual-exciting point process that captures both the temporal and semantic aspects of online activities. Our best model achieved an average F1 score of $0.77 \pm 0.04$ and an AUC ROC of $0.81 \pm 0.02$.

cs.CY↗
arXiv2020

The Relationship between Deteriorating Mental Health Conditions and Longitudinal Behavioral Changes in Google and YouTube Usages among College Students in the United States during COVID-19: Observational Study

Mental health problems among the global population are worsened during the coronavirus disease (COVID-19). How individuals engage with online platforms such as Google Search and YouTube undergoes drastic shifts due to pandemic and subsequent lockdowns. Such ubiquitous daily behaviors on online platforms have the potential to capture and correlate with clinically alarming deteriorations in mental health profiles in a non-invasive manner. The goal of this study is to examine, among college students, the relationship between deteriorating mental health conditions and changes in user behaviors when engaging with Google Search and YouTube during COVID-19. This study recruited a cohort of 49 students from a U.S. college campus during January 2020 (prior to the pandemic) and measured the anxiety and depression levels of each participant. This study followed up with the same cohort during May 2020 (during the pandemic), and the anxiety and depression levels were assessed again. The longitudinal Google Search and YouTube history data were anonymized and collected. From individual-level Google Search and YouTube histories, we developed 5 signals that can quantify shifts in online behaviors during the pandemic. We then assessed the differences between groups with and without deteriorating mental health profiles in terms of these features. Significant features included late-night online activities, continuous usages, and time away from the internet, porn consumptions, and keywords associated with negative emotions, social activities, and personal affairs. Though further studies are required, our results demonstrated the feasibility of utilizing pervasive online data to establish non-invasive surveillance systems for mental health conditions that bypasses many disadvantages of existing screening methods.

cs.CY↗
arXiv2020

A Novel DNN Training Framework via Data Sampling and Multi-Task Optimization

Conventional DNN training paradigms typically rely on one training set and one validation set, obtained by partitioning an annotated dataset used for training, namely gross training set, in a certain way. The training set is used for training the model while the validation set is used to estimate the generalization performance of the trained model as the training proceeds to avoid over-fitting. There exist two major issues in this paradigm. Firstly, the validation set may hardly guarantee an unbiased estimate of generalization performance due to potential mismatching with test data. Secondly, training a DNN corresponds to solve a complex optimization problem, which is prone to getting trapped into inferior local optima and thus leads to undesired training results. To address these issues, we propose a novel DNN training framework. It generates multiple pairs of training and validation sets from the gross training set via random splitting, trains a DNN model of a pre-specified structure on each pair while making the useful knowledge (e.g., promising network parameters) obtained from one model training process to be transferred to other model training processes via multi-task optimization, and outputs the best, among all trained models, which has the overall best performance across the validation sets from all pairs. The knowledge transfer mechanism featured in this new framework can not only enhance training effectiveness by helping the model training process to escape from local optima but also improve on generalization performance via implicit regularization imposed on one model training process from other model training processes. We implement the proposed framework, parallelize the implementation on a GPU cluster, and apply it to train several widely used DNN models. Experimental results demonstrate the superiority of the proposed framework over the conventional training paradigm.

cs.NE↗
arXiv2020

Instantons and Khovanov skein homology on $I\times T^2$

Asaeda, Przytycki and Sikora defined a generalization of Khovanov homology for links in $I$-bundles over compact surfaces. We prove that for a link $L\subset (-1,1)\times T^2$, the Asaeda-Przytycki-Sikora homology of $L$ has rank $2$ with $\mathbb{Z}/2$-coefficients if and only if $L$ is isotopic to an embedded knot in $\{0\}\times T^2$.

math.GT↗
arXiv2020

On links with Khovanov homology of small ranks

We classify all links whose Khovanov homology have ranks no greater than 8, and all three-component links whose Khovanov homology have ranks no greater than 12, where the coefficient ring is Z/2. The classification is based on the previous results of Kronheimer-Mrowka, Batson-Seed, Baldwin-Sivek, and the authors.

math.GT↗