arXiv subjects

Yilong Yang

Publications and source records attributed to Yilong Yang.

At least 37 records · Page 2Linked to original sources

arXiv2023

Scale-Equivariant UNet for Histopathology Image Segmentation

Digital histopathology slides are scanned and viewed under different magnifications and stored as images at different resolutions. Convolutional Neural Networks (CNNs) trained on such images at a given scale fail to generalise to those at different scales. This inability is often addressed by augmenting training data with re-scaled images, allowing a model with sufficient capacity to learn the requisite patterns. Alternatively, designing CNN filters to be scale-equivariant frees up model capacity to learn discriminative features. In this paper, we propose the Scale-Equivariant UNet (SEUNet) for image segmentation by building on scale-space theory. The SEUNet contains groups of filters that are linear combinations of Gaussian basis filters, whose scale parameters are trainable but constrained to span disjoint scales through the layers of the network. Extensive experiments on a nuclei segmentation dataset and a tissue type segmentation dataset demonstrate that our method outperforms other approaches, with much fewer trainable parameters.

cs.CV↗

arXiv2023

Rotation-Scale Equivariant Steerable Filters

Incorporating either rotation equivariance or scale equivariance into CNNs has proved to be effective in improving models' generalization performance. However, jointly integrating rotation and scale equivariance into CNNs has not been widely explored. Digital histology imaging of biopsy tissue can be captured at arbitrary orientation and magnification and stored at different resolutions, resulting in cells appearing in different scales. When conventional CNNs are applied to histopathology image analysis, the generalization performance of models is limited because 1) a part of the parameters of filters are trained to fit rotation transformation, thus decreasing the capability of learning other discriminative features; 2) fixed-size filters trained on images at a given scale fail to generalize to those at different scales. To deal with these issues, we propose the Rotation-Scale Equivariant Steerable Filter (RSESF), which incorporates steerable filters and scale-space theory. The RSESF contains copies of filters that are linear combinations of Gaussian filters, whose direction is controlled by directional derivatives and whose scale parameters are trainable but constrained to span disjoint scales in successive layers of the network. Extensive experiments on two gland segmentation datasets demonstrate that our method outperforms other approaches, with much fewer trainable parameters and fewer GPU resources required. The source code is available at: https://github.com/ynulonger/RSESF.

cs.CV↗

arXiv2023

Deep-neural-network approach to solving the ab initio nuclear structure problem

Predicting the structure of quantum many-body systems from the first principles of quantum mechanics is a common challenge in physics, chemistry, and material science. Deep machine learning has proven to be a powerful tool for solving condensed matter and chemistry problems, while for atomic nuclei it is still quite challenging because of the complicated nucleon-nucleon interactions, which strongly couple the spatial, spin, and isospin degrees of freedom. By combining essential physics of the nuclear wave functions and the strong expressive power of artificial neural networks, we develop FeynmanNet, a deep-learning variational quantum Monte Carlo approach for \emph{ab initio} nuclear structure. We show that FeynmanNet can provide very accurate solutions of ground-state energies and wave functions for $^4$He, $^6$Li, and even up to $^{16}$O as emerging from the leading-order and next-to-leading-order Hamiltonians of pionless effective field theory. Compared to the conventional diffusion Monte Carlo approaches, which suffer from the severe inherent fermion-sign problem, FeynmanNet reaches such a high accuracy in a variational way and scales polynomially with the number of nucleons. Therefore, it paves the way to a highly accurate and efficient \emph{ab initio} method for predicting nuclear properties based on the realistic interactions between nucleons.

nucl-th↗

arXiv2022

Reduced Power Graphs of $\mathrm{PGL}_n(\mathbb{F}_q)$

Given a group $G$, let us connect two non-identity elements by an edge if and only if one is a power of another. This gives a graph structure on $G$ minus identity, called the reduced power graph. It is conjectured by Akbari and Ashrafi that if a non-abelian finite simple group has a connected reduced power graph, then it must be an alternating group. In this paper, we shall give a complete description of when the reduced power graphs of $\mathrm{PGL}_n(\mathbb{F}_q)$ are connected for all $q$ and all $n\geq 3$. In particular, the conjectured by Akbari and Ashrafi is false. We shall also provide an upper bound in their diameters, and in case of disconnection, provide a description of all connected components.

math.GR↗

arXiv2022

Automated Enterprise Applications Generation from Requirements Model

Enterprise applications can be automatically generated from a sophisticated OO design model based on model-driven approach. The design model contains information about how to decompose the system into components, how to encapsulate the system operations into classes, and how the objects of classes collaborate to fulfill the functionality of the system operations. However, the efforts to build the design model from a validated requirements model are not proportional to the return. In practice, it is very desirable to have an approach that can automatically generate standardized enterprise applications directly from the validated requirements models. In this paper, we propose an approach named RM2EA, which can reach this goal based on the contract-based requirements model. We demonstrate the proposed approach through 13 case studies. The evaluation result shows that the quality and efficiency of the generated applications are almost equal to the applications implemented by developers: firstly, we demonstrate that a popular type of enterprise applications (i.e., a Jakarta EE application) can be successfully generated by customizing and improving the set of rules; secondly, RM2EA can generate more readable or maintainable code; thirdly, the enterprise applications generated by RM2EA achieve similar performance in test results. Overall, the result is satisfactory, and the implementation of the proposed approach can be further enhanced and applied to software development in the industry.

cs.SE↗

arXiv2021

Projective Limits and Ultraproducts of Nonabelian Finite Groups

Groups that can be approximated by finite groups have been the center of much research. This has led to the investigations of the subgroups of metric ultraproducts of finite groups. This paper attempts to study the dual problem: what are the quotients of ultraproducts of finite groups? Since an ultraproduct is an abstract quotient of the direct product, this also led to a more general question: what are the abstract quotients of profinite groups? Certain cases were already well-studied. For example, if we have a pro-solvable group, then it is well-known that any finite abstract quotients are still solvable. This paper studies the case of profinite groups from a surjective inverse system of non-solvable finite groups. We shall show that, for various classes of groups $\mathcal{C}$, any finite abstract quotient of a pro-$\mathcal{C}$ group is still in $\mathcal{C}$. Here $\mathcal{C}$ can be the class of finite semisimple groups, or the class of central products of quasisimple groups, or the class of products of finite almost simple groups and finite solvable groups, or the class of finite perfect groups with bounded commutator width. Furthermore, in the case of finite perfect groups with unbounded commutator width, we show that this is NOT the case. In fact, any finite group could be a quotient of an ultraproduct of finite perfect groups. We provide an explicit construction on how to achieve this for each finite group.

math.GR↗

arXiv2020

ServeNet: A Deep Neural Network for Web Services Classification

Automated service classification plays a crucial role in service discovery, selection, and composition. Machine learning has been widely used for service classification in recent years. However, the performance of conventional machine learning methods highly depends on the quality of manual feature engineering. In this paper, we present a novel deep neural network to automatically abstract low-level representation of both service name and service description to high-level merged features without feature engineering and the length limitation, and then predict service classification on 50 service categories. To demonstrate the effectiveness of our approach, we conduct a comprehensive experimental study by comparing 10 machine learning methods on 10,000 real-world web services. The result shows that the proposed deep neural network can achieve higher accuracy in classification and more robust than other machine learning methods.

cs.LG↗

arXiv2020

Cloud-based Federated Boosting for Mobile Crowdsensing

The application of federated extreme gradient boosting to mobile crowdsensing apps brings several benefits, in particular high performance on efficiency and classification. However, it also brings a new challenge for data and model privacy protection. Besides it being vulnerable to Generative Adversarial Network (GAN) based user data reconstruction attack, there is not the existing architecture that considers how to preserve model privacy. In this paper, we propose a secret sharing based federated learning architecture FedXGB to achieve the privacy-preserving extreme gradient boosting for mobile crowdsensing. Specifically, we first build a secure classification and regression tree (CART) of XGBoost using secret sharing. Then, we propose a secure prediction protocol to protect the model privacy of XGBoost in mobile crowdsensing. We conduct a comprehensive theoretical analysis and extensive experiments to evaluate the security, effectiveness, and efficiency of FedXGB. The results indicate that FedXGB is secure against the honest-but-curious adversaries and attains less than 1% accuracy loss compared with the original XGBoost model.

cs.CR↗

arXiv2018

An Integrated Transfer Learning and Multitask Learning Approach for Pharmacokinetic Parameter Prediction

Background: Pharmacokinetic evaluation is one of the key processes in drug discovery and development. However, current absorption, distribution, metabolism, excretion prediction models still have limited accuracy. Aim: This study aims to construct an integrated transfer learning and multitask learning approach for developing quantitative structure-activity relationship models to predict four human pharmacokinetic parameters. Methods: A pharmacokinetic dataset included 1104 U.S. FDA approved small molecule drugs. The dataset included four human pharmacokinetic parameter subsets (oral bioavailability, plasma protein binding rate, apparent volume of distribution at steady-state and elimination half-life). The pre-trained model was trained on over 30 million bioactivity data. An integrated transfer learning and multitask learning approach was established to enhance the model generalization. Results: The pharmacokinetic dataset was split into three parts (60:20:20) for training, validation and test by the improved Maximum Dissimilarity algorithm with the representative initial set selection algorithm and the weighted distance function. The multitask learning techniques enhanced the model predictive ability. The integrated transfer learning and multitask learning model demonstrated the best accuracies, because deep neural networks have the general feature extraction ability, transfer learning and multitask learning improved the model generalization. Conclusions: The integrated transfer learning and multitask learning approach with the improved dataset splitting algorithm was firstly introduced to predict the pharmacokinetic parameters. This method can be further employed in drug discovery and development.

cs.LG↗

arXiv2018

Deep learning for in vitro prediction of pharmaceutical formulations

Current pharmaceutical formulation development still strongly relies on the traditional trial-and-error approach by individual experiences of pharmaceutical scientists, which is laborious, time-consuming and costly. Recently, deep learning has been widely applied in many challenging domains because of its important capability of automatic feature extraction. The aim of this research is to use deep learning to predict pharmaceutical formulations. In this paper, two different types of dosage forms were chosen as model systems. Evaluation criteria suitable for pharmaceutics were applied to assessing the performance of the models. Moreover, an automatic dataset selection algorithm was developed for selecting the representative data as validation and test datasets. Six machine learning methods were compared with deep learning. The result shows the accuracies of both two deep neural networks were above 80% and higher than other machine learning models, which showed good prediction in pharmaceutical formulations. In summary, deep learning with the automatic data splitting algorithm and the evaluation criteria suitable for pharmaceutical formulation data was firstly developed for the prediction of pharmaceutical formulations. The cross-disciplinary integration of pharmaceutics and artificial intelligence may shift the paradigm of pharmaceutical researches from experience-dependent studies to data-driven methodologies.

cs.LG↗

arXiv2018

Automated Prototype Generation from Formal Requirements Model

Prototyping is an effective and efficient way of requirement validation to avoid introducing errors in the early stage of software development. However, manually developing a prototype of a software system requires additional efforts, which would increase the overall cost of software development. In this paper, we present an approach with a developed tool to automatic generation of prototypes from formal requirements models. A requirements model consists of a use case diagram, a conceptual class diagram, use case definitions specified by system sequence diagrams and the contracts of their system operations. We propose a method to decompose a contract into executable parts and non-executable parts. A set of transformation rules is given to decompose the executable part into pre-implemented primitive operations. A non-executable part is usually realized by significant algorithms such as sorting a list, finding the shortest path or domain-specific computation. It can be implemented manually or by using existing code. A CASE tool is developed that provides an interface for developers to develop a program for each non-executable part of a contract, and automatically transforms the executables into sequences of pre-implemented primitive operations. We have conducted four cases studies with over 50 use cases. The experimental result shows that the 93.65% of requirement specifications are executable, and only 6.35% are non-executable such as sorting and event-call, which can be implemented by developers manually or invoking the APIs of advanced algorithms in Java library. The one second generated the prototype of a case study requires approximate nine hours manual implementation by a skilled programmer. Overall, the result is satisfiable, and the proposed approach with the developed CASE tool can be applied to the software industry for requirements engineering.

cs.SE↗

arXiv2018

MicroShare: Privacy-Preserved Medical Resource Sharing through MicroService Architecture

This paper takes up the problem of medical resource sharing through MicroService architecture without compromising patient privacy. To achieve this goal, we suggest refactoring the legacy EHR systems into autonomous MicroServices communicating by the unified techniques such as RESTFul web service. This lets us handle clinical data queries directly and far more efficiently for both internal and external queries. The novelty of the proposed approach lies in avoiding the data de-identification process often used as a means of preserving patient privacy. The implemented toolkit combines software engineering technologies such as Java EE, RESTful web services, JSON Web Tokens to allow exchanging medical data in an unidentifiable XML and JSON format as well as restricting users to the need-to-know principle. Our technique also inhibits retrospective processing of data such as attacks by an adversary on a medical dataset using advanced computational methods to reveal Protected Health Information (PHI). The approach is validated on an endoscopic reporting application based on openEHR and MST standards. From the usability perspective, the approach can be used to query datasets by clinical researchers, governmental or non-governmental organizations in monitoring health care and medical record services to improve quality of care and treatment.

cs.SE↗

arXiv2018

Predicting Oral Disintegrating Tablet Formulations by Neural Network Techniques

Oral Disintegrating Tablets (ODTs) is a novel dosage form that can be dissolved on the tongue within 3min or less especially for geriatric and pediatric patients. Current ODT formulation studies usually rely on the personal experience of pharmaceutical experts and trial-and-error in the laboratory, which is inefficient and time-consuming. The aim of current research was to establish the prediction model of ODT formulations with direct compression process by Artificial Neural Network (ANN) and Deep Neural Network (DNN) techniques. 145 formulation data were extracted from Web of Science. All data sets were divided into three parts: training set (105 data), validation set (20) and testing set (20). ANN and DNN were compared for the prediction of the disintegrating time. The accuracy of the ANN model has reached 85.60%, 80.00% and 75.00% on the training set, validation set and testing set respectively, whereas that of the DNN model was 85.60%, 85.00% and 80.00%, respectively. Compared with the ANN, DNN showed the better prediction for ODT formulations. It is the first time that deep neural network with the improved dataset selection algorithm is applied to formulation prediction on small data. The proposed predictive approach could evaluate the critical parameters about quality control of formulation, and guide research and process development. The implementation of this prediction model could effectively reduce drug product development timeline and material usage, and proactively facilitate the development of a robust drug product.

stat.ML↗

arXiv2018

MedShare: Medical Resource Sharing among Autonomous Healthcare Providers

Legacy Electronic Health Records (EHRs) systems were not developed with the level of connectivity expected from them nowadays. Therefore, interoperability weakness inherent in the legacy systems can result in poor patient care and waste of financial resources. Large hospitals are less likely to share their data with external hospitals due to economic and political reasons. Motivated by these facts, we aim to provide a set of software implementation guidelines, i.e., MedShare to deal with interoperability issues among disconnected healthcare systems. The proposed integrated architecture includes: 1) a data extractor to fetch legacy medical data from a hemodialysis center, 2) converting it to a common data model, 3) indexing patient information using the HashMap technique, and 4) a set of services and tools that can be installed as a coherent environment on top of stand-alone EHRs systems. Our work enabled three cooperating but autonomous hospitals to mutually exchange medical data and helped them develop a common reference architecture. It lets stakeholders retain control over their patient data, winning the trust and confidence much needed towards a successful deployment of MedShare. Security concerns were effectively addressed that also included patient consent in the data exchange process. Thereby, the implemented toolset offered a collaborative environment to share EHRs by the healthcare providers.

cs.SE↗

arXiv2018

Integrating UML with Service Refinement for Requirements Modeling and Analysis

Unified Modeling Language (UML) is the de facto standard for requirements modeling and system design. UML as a visual language can tremendously help customers, project managers, and developers to specify the requirements of a target system. However, UML lacks the ability to specify the requirements precisely such as the contracts of the system operation, and verify the consistency and refinement of the requirements. These disadvantages result in that the potential faults of software are hard to be discovered in the early stage of software development process, and then requiring more efforts in software testing to find the bugs. Service refinement is a formal method, which could be a supplement to enhance the UML. In this paper, we show how to integrate UML with service refinement to specify requirements, and verify the consistency and refinements of the requirements through a case study of online shopping system. Particularly, requirements are modeled through UML diagrams, which includes a) use case diagram, b) system sequence diagrams and c) conceptual class diagram. Service refinement enhances the requirements model by introducing the contracts. Furthermore, the consistency and refinements of requirement model can be verified through service refinement. Our approach demonstrates integrating UML with service refinement can require fewer efforts to achieve the consistency requirements than only using UML for requirement modeling.

cs.SE↗

arXiv2018

Real-time System Modeling and Verification through Labeled Transition System Analyser (LTSA)

With the advancement of software engineering in recent years, the model checking techniques are widely applied in various areas to do the verification for the system model. However, it is difficult to apply the model checking to verify requirements due to lacking the details of the design. Unlike other model checking tools, LTSA provides the structure diagram, which can bridge the gap between the requirements and the design. In this paper, we demonstrate the abilities of LTSA shipped with the classic case study of the steam boiler system. The structure diagram of LTSA can specify the interactions between the controller and the steam boiler, which can be derived from UML requirements model such as system sequence diagram of the steam boiler system. The start-up design model of LTSA can be generated from the structure diagram. Furthermore, we provide a variation law of the steam rate to avoid the issue of state space explosion and show how explicitly and implicitly model the time that reflects the difference between system modeling and the physical world. Finally, the derived model is verified against the required properties. Our work demonstrates the potential power of integrating UML with model checking tools in requirement elicitation, system design, and verification.

cs.SE↗

arXiv2016

The Ultraproducts of Quasirandom Groups

In this paper, we shall prove that an ultraproduct of non-abelian finite simple groups is either finite simple, or has no finite dimensional unitary representation other than the trivial one. Then we shall generalize this result for other kinds of quasirandom groups. A group is called D- quasirandom if all of its nontrivial representations over the complex numbers have dimensions at least D. We shall study the question of whether a non-principal ultraproduct of a given sequence of quasirandom groups remains quasirandom, and whether an ultraproduct of increasingly quasirandom groups becomes minimally almost periodic (i.e. no non-trivial finite-dimensional unitary representation at all). We answer this question in the affirmative when the groups in question are simple, quasisimple, semisimple, or when the groups in question have bounded number of conjugacy classes in their cosocles (the intersection of all maximal normal subgroups), or when the groups are arbitrary products (not necessarily finite) of the groups just listed. We shall also present with an ultraproduct of increasingly quasirandom groups with a non-trivial one-dimensional representation. We also obtain some results in the case of semi-direct products and short exact sequences of quasirandom groups. Finally, two applications of our results are given, one in triangle patterns of quasirandom groups and one in self-Bohrifying groups. Our main tools are some variations of the covering number for groups, different kinds of length functions on groups, and the classification of finite simple groups.

math.GR↗

arXiv2016

Billiards in near rectangles

For every quadrilateral sufficiently close to a rectangle, we shall show that it possess a periodic billiard path. This is an REU work done at ICERM in Summer 2012.

math.DS↗