arXiv subjects

Xiaolin Li

Publications and source records attributed to Xiaolin Li.

At least 55 records · Page 3Linked to original sources

arXiv2020

Improving Question Generation with Sentence-level Semantic Matching and Answer Position Inferring

Taking an answer and its context as input, sequence-to-sequence models have made considerable progress on question generation. However, we observe that these approaches often generate wrong question words or keywords and copy answer-irrelevant words from the input. We believe that lacking global question semantics and exploiting answer position-awareness not well are the key root causes. In this paper, we propose a neural question generation model with two concrete modules: sentence-level semantic matching and answer position inferring. Further, we enhance the initial state of the decoder by leveraging the answer-aware gated fusion mechanism. Experimental results demonstrate that our model outperforms the state-of-the-art (SOTA) models on SQuAD and MARCO datasets. Owing to its generality, our work also improves the existing models significantly.

cs.CL↗

arXiv2019

DeepAtom: A Framework for Protein-Ligand Binding Affinity Prediction

The cornerstone of computational drug design is the calculation of binding affinity between two biological counterparts, especially a chemical compound, i.e., a ligand, and a protein. Predicting the strength of protein-ligand binding with reasonable accuracy is critical for drug discovery. In this paper, we propose a data-driven framework named DeepAtom to accurately predict the protein-ligand binding affinity. With 3D Convolutional Neural Network (3D-CNN) architecture, DeepAtom could automatically extract binding related atomic interaction patterns from the voxelized complex structure. Compared with the other CNN based approaches, our light-weight model design effectively improves the model representational capacity, even with the limited available training data. With validation experiments on the PDBbind v.2016 benchmark and the independent Astex Diverse Set, we demonstrate that the less feature engineering dependent DeepAtom approach consistently outperforms the other state-of-the-art scoring methods. We also compile and propose a new benchmark dataset to further improve the model performances. With the new dataset as training input, DeepAtom achieves Pearson's R=0.83 and RMSE=1.23 pK units on the PDBbind v.2016 core set. The promising results demonstrate that DeepAtom models can be potentially adopted in computational drug development protocols such as molecular docking and virtual screening.

q-bio.BM↗

arXiv2019

Adaptive Leader-Follower Formation Control and Obstacle Avoidance via Deep Reinforcement Learning

We propose a deep reinforcement learning (DRL) methodology for the tracking, obstacle avoidance, and formation control of nonholonomic robots. By separating vision-based control into a perception module and a controller module, we can train a DRL agent without sophisticated physics or 3D modeling. In addition, the modular framework averts daunting retrains of an image-to-action end-to-end neural network, and provides flexibility in transferring the controller to different robots. First, we train a convolutional neural network (CNN) to accurately localize in an indoor setting with dynamic foreground/background. Then, we design a new DRL algorithm named Momentum Policy Gradient (MPG) for continuous control tasks and prove its convergence. We also show that MPG is robust at tracking varying leader movements and can naturally be extended to problems of formation control. Leveraging reward shaping, features such as collision and obstacle avoidance can be easily integrated into a DRL controller.

cs.RO↗

arXiv2019

Added Value of Intraoperative Data for Predicting Postoperative Complications: Development and Validation of a MySurgeryRisk Extension

To test the hypothesis that accuracy, discrimination, and precision in predicting postoperative complications improve when using both preoperative and intraoperative data input features versus preoperative data alone. Models that predict postoperative complications often ignore important intraoperative physiological changes. Incorporation of intraoperative physiological data may improve model performance. This retrospective cohort analysis included 52,529 inpatient surgeries at a single institution during a 5 year period. Random forest machine learning models in the validated MySurgeryRisk platform made patient-level predictions for three postoperative complications and mortality during hospital admission using electronic health record data and patient neighborhood characteristics. For each outcome, one model trained with preoperative data alone and one model trained with both preoperative and intraoperative data. Models were compared by accuracy, discrimination (expressed as AUROC), precision (expressed as AUPRC), and reclassification indices (NRI). Machine learning models incorporating both preoperative and intraoperative data had greater accuracy, discrimination, and precision than models using preoperative data alone for predicting all three postoperative complications (intensive care unit length of stay >48 hours, mechanical ventilation >48 hours, and neurological complications including delirium) and in-hospital mortality (accuracy: 88% vs. 77%, AUROC: 0.93 vs. 0.87, AUPRC: 0.21 vs. 0.15). Overall reclassification improvement was 2.9-10.0% for complications and 11.2% for in-hospital mortality. Incorporating both preoperative and intraoperative data significantly increased accuracy, discrimination, and precision for machine learning models predicting postoperative complications.

cs.CY↗

arXiv2019

Identifying MgII Narrow Absorption Lines with Deep Learning

Metal absorption line systems in distant quasar spectra probe of the history of gas content in the universe. The MgII $λλ$ 2796, 2803 doublet is one of the most important absorption lines since it is a proxy of the star formation rate and a tracer of the cold gas associated with high redshift galaxies. Machine learning algorithms have been used to detect absorption lines systems in large sky surveys, such as Principle Component Analysis (PCA), Gaussian Process (GP) and decision trees. A very powerful algorithm in the field of machine learning called deep neural networks, or '' deep learning'' is a new structure of neural network that automatically extracts semantic features from raw data and represents them at a high level. In this paper, we apply a deep convolutional neural network for absorption line detection. We use the previously published DR7 MgII catalog (Zhu et al. 2013) as the training and validation sample and the DR12 MgII catalog as the test set. Our deep learning algorithm is capable of detecting MgII absorption lines with an accuracy of $\sim$94% . It takes only $\sim 9$ seconds to analyze $\sim$ 50000 quasar spectra with our deep neural network, which is ten thousand times faster than traditional methods, while preserving high accuracy with little human interference. Our study shows that Mg II absorption line detection accuracy of a deep neutral network model strongly depends on the filter size in the filter layer of the neural network, and the best results are obtained when the filter size closely matches the absorption feature size.

astro-ph.GA↗

arXiv2019

Microstructural Materials Design via Deep Adversarial Learning Methodology

Identifying the key microstructure representations is crucial for Computational Materials Design (CMD). However, existing microstructure characterization and reconstruction (MCR) techniques have limitations to be applied for materials design. Model-based MCR approaches do not have parameters that can serve as design variables, while MCR techniques that rely on dimension reduction tend to lose important microstructural information. In this work, we present a deep adversarial learning methodology that overcomes the limitations of existing MCR techniques. In the proposed methodology, generative adversarial networks (GAN) are trained to learn the mapping between latent variables and microstructures. Thereafter, the low-dimensional latent variables serve as design variables, and a Bayesian optimization framework is applied to obtain microstructures with desired material property. Due to the special design of the network architecture, the proposed methodology is able to identify the latent (design) variables with desired dimensionality, as well as minimize the information loss even for complex material microstructures. The validity of the proposed methodology is tested numerically on a synthetic microstructure dataset and its effectiveness for materials design is evaluated through a case study of optimizing optical performance for energy absorption. In addition, the scalability and transferability of proposed methodology are also demonstrated in this work. Specifically, the proposed methodology is scalable to generate arbitrary sized microstructures, and it can serve as a pre-trained model to improve the performance of a structure-property predictive model via transfer learning.

cond-mat.mtrl-sci↗

arXiv2018

Generalized Batch Normalization: Towards Accelerating Deep Neural Networks

Utilizing recently introduced concepts from statistics and quantitative risk management, we present a general variant of Batch Normalization (BN) that offers accelerated convergence of Neural Network training compared to conventional BN. In general, we show that mean and standard deviation are not always the most appropriate choice for the centering and scaling procedure within the BN transformation, particularly if ReLU follows the normalization step. We present a Generalized Batch Normalization (GBN) transformation, which can utilize a variety of alternative deviation measures for scaling and statistics for centering, choices which naturally arise from the theory of generalized deviation measures and risk theory in general. When used in conjunction with the ReLU non-linearity, the underlying risk theory suggests natural, arguably optimal choices for the deviation measure and statistic. Utilizing the suggested deviation measure and statistic, we show experimentally that training is accelerated more so than with conventional BN, often with improved error rate as well. Overall, we propose a more flexible BN transformation supported by a complimentary theoretical framework that can potentially guide design choices.

cs.LG↗

arXiv2018

FoldingZero: Protein Folding from Scratch in Hydrophobic-Polar Model

De novo protein structure prediction from amino acid sequence is one of the most challenging problems in computational biology. As one of the extensively explored mathematical models for protein folding, Hydrophobic-Polar (HP) model enables thorough investigation of protein structure formation and evolution. Although HP model discretizes the conformational space and simplifies the folding energy function, it has been proven to be an NP-complete problem. In this paper, we propose a novel protein folding framework FoldingZero, self-folding a de novo protein 2D HP structure from scratch based on deep reinforcement learning. FoldingZero features the coupled approach of a two-head (policy and value heads) deep convolutional neural network (HPNet) and a regularized Upper Confidence Bounds for Trees (R-UCT). It is trained solely by a reinforcement learning algorithm, which improves HPNet and R-UCT iteratively through iterative policy optimization. Without any supervision and domain knowledge, FoldingZero not only achieves comparable results, but also learns the latent folding knowledge to stabilize the structure. Without exponential computation, FoldingZero shows promising potential to be adopted for real-world protein properties prediction.

cs.LG↗

arXiv2018

Rethinking Interphase Representations for Modeling Viscoelastic Properties for Polymer Nanocomposites

Numerical modeling of viscoelastic properties is critical to developing the structure-property relationship of polymer nanocomposites. While it is recognized that the altered polymer region near filler particles, the interphase, significantly contributes to enhancements of composite properties, the spatial distribution of interphase properties is rarely considered due to lack of local property measurements. In recent years, the Atomic Force Microscopy (AFM) technique has begun to make local property measurements of the interphase available. In the light of the increasing availability of AFM data, in this work a new interphase representation for modeling the viscoelastic properties of polymer nanocomposites is proposed. The proposed interphase representation disentangles the interphase behavior by two separate components -- single-body interphase gradient and multi-body compound effect, whose functional forms are learned via data mining. The proposed interphase representation is integrated into a Finite Element model, and the effects of each component of the interphase representations are numerically studied. In addition, the advantages of the proposed interphase representation are demonstrated by comparison to a prior simulation work in which only a uniform effective property of the interphase is considered. Moreover, the proposed interphase representation is utilized to solve the inverse problem of inferring spatial distribution of local properties using Bayesian Inference on experimental data.

cond-mat.mtrl-sci↗

arXiv2018

Adversarial Examples: Attacks and Defenses for Deep Learning

With rapid progress and significant successes in a wide spectrum of applications, deep learning is being applied in many safety-critical environments. However, deep neural networks have been recently found vulnerable to well-designed input samples, called adversarial examples. Adversarial examples are imperceptible to human but can easily fool deep neural networks in the testing/deploying stage. The vulnerability to adversarial examples becomes one of the major risks for applying deep neural networks in safety-critical environments. Therefore, attacks and defenses on adversarial examples draw great attention. In this paper, we review recent findings on adversarial examples for deep neural networks, summarize the methods for generating adversarial examples, and propose a taxonomy of these methods. Under the taxonomy, applications for adversarial examples are investigated. We further elaborate on countermeasures for adversarial examples and explore the challenges and the potential solutions.

cs.LG↗

arXiv2018

Improved Predictive Models for Acute Kidney Injury with IDEAs: Intraoperative Data Embedded Analytics

Acute kidney injury (AKI) is a common and serious complication after a surgery which is associated with morbidity and mortality. The majority of existing perioperative AKI risk score prediction models are limited in their generalizability and do not fully utilize the physiological intraoperative time-series data. Thus, there is a need for intelligent, accurate, and robust systems, able to leverage information from large-scale data to predict patient's risk of developing postoperative AKI. A retrospective single-center cohort of 2,911 adult patients who underwent surgery at the University of Florida Health has been used for this study. We used machine learning and statistical analysis techniques to develop perioperative models to predict the risk of AKI (risk during the first 3 days, 7 days, and until the discharge day) before and after the surgery. In particular, we examined the improvement in risk prediction by incorporating three intraoperative physiologic time series data, i.e., mean arterial blood pressure, minimum alveolar concentration, and heart rate. For an individual patient, the preoperative model produces a probabilistic AKI risk score, which will be enriched by integrating intraoperative statistical features through a machine learning stacking approach inside a random forest classifier. We compared the performance of our model based on the area under the receiver operating characteristics curve (AUROC), accuracy and net reclassification improvement (NRI). The predictive performance of the proposed model is better than the preoperative data only model. For AKI-7day outcome: The AUC was 0.86 (accuracy was 0.78) in the proposed model, while the preoperative AUC was 0.84 (accuracy 0.76). Furthermore, with the integration of intraoperative features, we were able to classify patients who were misclassified in the preoperative model.

cs.CY↗

arXiv2018

A Transfer Learning Approach for Microstructure Reconstruction and Structure-property Predictions

Stochastic microstructure reconstruction has become an indispensable part of computational materials science, but ongoing developments are specific to particular material systems. In this paper, we address this generality problem by presenting a transfer learning-based approach for microstructure reconstruction and structure-property predictions that is applicable to a wide range of material systems. The proposed approach incorporates an encoder-decoder process and feature-matching optimization using a deep convolutional network. For microstructure reconstruction, model pruning is implemented in order to study the correlation between the microstructural features and hierarchical layers within the deep convolutional network. Knowledge obtained in model pruning is then leveraged in the development of a structure-property predictive model to determine the network architecture and initialization conditions. The generality of the approach is demonstrated numerically for a wide range of material microstructures with geometrical characteristics of varying complexity. Unlike previous approaches that only apply to specific material systems or require a significant amount of prior knowledge in model selection and hyper-parameter tuning, the present approach provides an off-the-shelf solution to handle complex microstructures, and has the potential of expediting the discovery of new materials.

cond-mat.mtrl-sci↗

arXiv2018

Comparing Clinical Judgment with MySurgeryRisk Algorithm for Preoperative Risk Assessment: A Pilot Study

Background: Major postoperative complications are associated with increased short and long-term mortality, increased healthcare cost, and adverse long-term consequences. The large amount of data contained in the electronic health record (EHR) creates barriers for physicians to recognize patients most at risk. We hypothesize, if presented in an optimal format, information from data-driven predictive risk algorithms for postoperative complications can improve physician risk assessment. Methods: Prospective, non-randomized, interventional pilot study of twenty perioperative physicians at a quarterly academic medical center. Using 150 clinical cases we compared physicians' risk assessment before and after interaction with MySurgeryRisk, a validated machine-learning algorithm predicting preoperative risk for six major postoperative complications using EHR data. Results: The area under the curve (AUC) of MySurgeryRisk algorithm ranged between 0.73 and 0.85 and was significantly higher than physicians' risk assessments (AUC between 0.47 and 0.69) for all postoperative complications except cardiovascular complications. The AUC for repeated physician's risk assessment improved by 2% to 5% for all complications with the exception of thirty-day mortality. Physicians' risk assessment for acute kidney injury and intensive care unit admission longer than 48 hours significantly improved after knowledge exchange, resulting in net reclassification improvement of 12.4% and 16%, respectively. Conclusions: The validated MySurgeryRisk algorithm predicted postoperative complications with equal or higher accuracy than pilot cohort of physicians using available clinical preoperative data. The interaction with algorithm significantly improved physicians' risk assessment.

cs.HC↗

arXiv2018

End-to-End Service Delivery with QoS Guarantee in Software Defined Networks

Software-Defined Network (SDN) is expected to have a significant impact on future networking. Although exciting progress has been made toward realizing SDN, application of this new networking paradigm in the future Internet to support end-to-end QoS provisioning faces some new challenges. The autonomous network domains coexisting in the Internet and the diverse user applications deployed upon the Internet call for a uniform Service Delivery Platform (SDP) that enables high-level network abstraction and inter-domain collaboration for end-to-end service provisioning. However, the currently available SDN technologies lack effective mechanisms for supporting such a platform. In this paper, we first present a SDP framework that applies the Network-as-a-Service (NaaS) principle to provide network abstraction and orchestration for end-to-end service provisioning in SDN-based future Internet. Then we focus our study on two enabling technologies for such a SDP to achieve QoS guarantee; namely a network abstraction model and an end-to-end resource allocation scheme. Specifically we propose a general model for abstracting the service capabilities offered by network domains and develop a technique for determining the required amounts of bandwidth in network domains for end-to-end service delivery with QoS guarantee. Both the analytical and numerical results obtained in this paper indicate that the NaaS-based SDP not only simplifies SDN service and resource management but also enhances bandwidth utilization for end-to-end QoS provisioning.

cs.NI↗

arXiv2017

Intelligent Perioperative System: Towards Real-time Big Data Analytics in Surgery Risk Assessment

Surgery risk assessment is an effective tool for physicians to manage the treatment of patients, but most current research projects fall short in providing a comprehensive platform to evaluate the patients' surgery risk in terms of different complications. The recent evolution of big data analysis techniques makes it possible to develop a real-time platform to dynamically analyze the surgery risk from large-scale patients information. In this paper, we propose the Intelligent Perioperative System (IPS), a real-time system that assesses the risk of postoperative complications (PC) and dynamically interacts with physicians to improve the predictive results. In order to process large volume patients data in real-time, we design the system by integrating several big data computing and storage frameworks with the high through-output streaming data processing components. We also implement a system prototype along with the visualization results to show the feasibility of system design.

cs.SE↗

arXiv2017

Single Shot Text Detector with Regional Attention

We present a novel single-shot text detector that directly outputs word-level bounding boxes in a natural image. We propose an attention mechanism which roughly identifies text regions via an automatically learned attentional map. This substantially suppresses background interference in the convolutional features, which is the key to producing accurate inference of words, particularly at extremely small sizes. This results in a single model that essentially works in a coarse-to-fine manner. It departs from recent FCN- based text detectors which cascade multiple FCN models to achieve an accurate prediction. Furthermore, we develop a hierarchical inception module which efficiently aggregates multi-scale inception features. This enhances local details, and also encodes strong context information, allow- ing the detector to work reliably on multi-scale and multi- orientation text with single-scale images. Our text detector achieves an F-measure of 77% on the ICDAR 2015 bench- mark, advancing the state-of-the-art results in [18, 28]. Demo is available at: http://sstd.whuang.org/.

cs.CV↗

arXiv2016

DeepCancer: Detecting Cancer through Gene Expressions via Deep Generative Learning

Transcriptional profiling on microarrays to obtain gene expressions has been used to facilitate cancer diagnosis. We propose a deep generative machine learning architecture (called DeepCancer) that learn features from unlabeled microarray data. These models have been used in conjunction with conventional classifiers that perform classification of the tissue samples as either being cancerous or non-cancerous. The proposed model has been tested on two different clinical datasets. The evaluation demonstrates that DeepCancer model achieves a very high precision score, while significantly controlling the false positive and false negative scores.

cs.AI↗

arXiv2015

Identifying the Absorption Bump with Deep Learning

The pervasive interstellar dust grains provide significant insights to understand the formation and evolution of the stars, planetary systems, and the galaxies, and may harbor the building blocks of life. One of the most effective way to analyze the dust is via their interaction with the light from background sources. The observed extinction curves and spectral features carry the size and composition information of dust. The broad absorption bump at 2175 Angstrom is the most prominent feature in the extinction curves. Traditionally, statistical methods are applied to detect the existence of the absorption bump. These methods require heavy preprocessing and the co-existence of other reference features to alleviate the influence from the noises. In this paper, we apply Deep Learning techniques to detect the broad absorption bump. We demonstrate the key steps for training the selected models and their results. The success of Deep Learning based method inspires us to generalize a common methodology for broader science discovery problems. We present our on-going work to build the DeepDis system for such kind of applications.

cs.CV↗