Search

PURPOSE: Herpes zoster ophthalmicus (HZO) after COVID-19 vaccination has been reported in numerous case studies. However, no large-scale epidemiologic studies have been conducted to date. The purpose of this study was to determine whether COVID-19 vaccination is associated with an increased risk of HZO. DESIGN: Retrospective before-and-after risk interval analysis. METHODS: RESULTS: In total, 1,959,157 patients received a dose of a COVID-19 vaccine during the study period and met eligibility criteria. A total of 80 individuals without a prior history of HZO were included in the analysis because they developed HZO in the risk or control period. Patients had a mean age of 54.0 years (SD = 12.3 years). There were 45 cases of HZO in the risk interval after COVID-19 vaccination. There was not an increased risk of HZO after vaccination with BNT162b2 (IRR = 0.90, 95% CI: 0.49-1.69, P = .74), mRNA-1273 (IRR = 0.74, 95% CI: 0.36-1.54, P = .42), or Ad26.COV2.S (IRR = 0.50, 95% CI: 0.07-2.56, P = .42). CONCLUSIONS: This study found no evidence of increased risk of HZO after COVID-19 vaccination, providing reassurance for patients and providers who may be concerned about the safety profile of the COVID-19 vaccines.

Neural networks encounter the challenge of Catastrophic Forgetting (CF) in continual learning, where new task learning interferes with previously learned knowledge. Existing data fine-tuning and regularization methods necessitate task identity information during inference and cannot eliminate interference among different tasks, while soft parameter sharing approaches encounter the problem of an increasing model parameter size. To tackle these challenges, we propose the Remembering Transformer, inspired by the brain's Complementary Learning Systems (CLS). Remembering Transformer employs a mixture-of-adapters architecture and a generative model-based novelty detection mechanism in a pretrained Transformer to alleviate CF. Remembering Transformer dynamically routes task data to the most relevant adapter with enhanced parameter efficiency based on knowledge distillation. We conducted extensive experiments, including ablation studies on the novelty detection mechanism and model capacity of the mixture-of-adapters, in a broad range of class-incremental split tasks and permutation tasks. Our approach demonstrated SOTA performance surpassing the second-best method by 15.90% in the split tasks, reducing the memory footprint from 11.18M to 0.22M in the five splits CIFAR10 task.

Meta-learning usually refers to a learning algorithm that learns from other learning algorithms. The problem of uncertainty in the predictions of neural networks shows that the world is only partially predictable and a learned neural network cannot generalize to its ever-changing surrounding environments. Therefore, the question is how a predictive model can represent multiple predictions simultaneously. We aim to provide a fundamental understanding of learning to learn in the contents of Decentralized Neural Networks (Decentralized NNs) and we believe this is one of the most important questions and prerequisites to building an autonomous intelligence machine. To this end, we shall demonstrate several pieces of evidence for tackling the problems above with Meta Learning in Decentralized NNs. In particular, we will present three different approaches to building such a decentralized learning system: (1) learning from many replica neural networks, (2) building the hierarchy of neural networks for different functions, and (3) leveraging different modality experts to learn cross-modal representations., Comment: Accepted for AAAI 2023 Doctoral Consortium

Meta-learning aims to develop algorithms that can learn from other learning algorithms to adapt to new and changing environments. This requires a model of how other learning algorithms operate and perform in different contexts, which is similar to representing and reasoning about mental states in the theory of mind. Furthermore, the problem of uncertainty in the predictions of conventional deep neural networks highlights the partial predictability of the world, requiring the representation of multiple predictions simultaneously. This is facilitated by coordination among neural modules, where different modules' beliefs and desires are attributed to others. The neural coordination among modular and decentralized neural networks is a fundamental prerequisite for building autonomous intelligence machines that can interact flexibly and adaptively. In this work, several pieces of evidence demonstrate a new avenue for tackling the problems above, termed Meta Neural Coordination. We discuss the potential advancements required to build biologically-inspired machine intelligence, drawing from both machine learning and cognitive science communities.

Trojan attacks embed perturbations in input data leading to malicious behavior in neural network models. A combination of various Trojans in different modalities enables an adversary to mount a sophisticated attack on multimodal learning such as Visual Question Answering (VQA). However, multimodal Trojans in conventional methods are susceptible to parameter adjustment during processes such as fine-tuning. To this end, we propose an instance-level multimodal Trojan attack on VQA that efficiently adapts to fine-tuned models through a dual-modality adversarial learning method. This method compromises two specific neurons in a specific perturbation layer in the pretrained model to produce overly large neuron activations. Then, a malicious correlation between these overactive neurons and the malicious output of a fine-tuned model is established through adversarial learning. Extensive experiments are conducted using the VQA-v2 dataset, based on a wide range of metrics including sample efficiency, stealthiness, and robustness. The proposed attack demonstrates enhanced performance with diverse vision and text Trojans tailored for each sample. We demonstrate that the proposed attack can be efficiently adapted to different fine-tuned models, by injecting only a few shots of Trojan samples. Moreover, we investigate the attack performance under conventional defenses, where the defenses cannot effectively mitigate the attack., Comment: Accepted for IJCNN 2024

Recombinant zoster vaccine (RZV) is recommended for individuals ≥ 50 years of age for protection against herpes zoster (HZ). This study quantifies RZV coverage and assesses predictors for RZV vaccination using a U.S. claims database. Univariate linear regression provided annual prevalence of RZV vaccination and multivariable logistic regression provided ORs and 95% CIs for associations between predictors and RZV vaccination. A total of 4,124,315 individuals (19,080,914 person-years) were included in this study. Since receiving FDA approval for the prevention of HZ, RZV coverage (of at least one dose) has reached approximately 17% within the eligible U.S. population by January 2021, although significant disparities between demographic groups were noted. Our findings suggest that HZ vaccine coverage may be reduced below goal in the U.S. and highlights the importance of continuing to monitor RZV vaccination. Additionally, as our study found disparities in vaccine coverage, attention towards marginalized and medically underserved populations is needed.

PURPOSE: To assess noninfectious uveitis (NIU) risk after coronavirus disease 2019 (COVID-19) vaccination in patients without a history of uveitis. DESIGN: A retrospective matched cohort study and self-controlled case series (SCCS) analysis using a longitudinal data asset with claims data from the OptumLabs Data Warehouse from December 11, 2020, through November 30, 2021. PARTICIPANTS: The matched cohort analysis included patients continuously enrolled for 730 days before December 11, 2020, who received a COVID-19 vaccination during the study period. This COVID-19-vaccinated group was matched to a COVID-19-unvaccinated historical cohort enrolled in 2018 and 2019. The SCCS design included individuals from the vaccinated cohort who experienced an NIU event during the study period. Enrollees with a history of uveitis were excluded. METHODS: Hazard ratios (HRs) were calculated using Cox proportional hazards models in the matched cohort design. Incidence rate ratios (IRRs) comparing NIU incidence in exposed risk periods after vaccination and unexposed control periods within individuals were calculated using conditional Poisson regression models in the SCCS design. Models were adjusted for age, recent receipt of non-COVID-19 vaccinations, corticosteroid or immunosuppressive use, and smoking history. Subgroup analyses were conducted by vaccination type and age group. MAIN OUTCOME MEASURES: Rates of NIU identified with International Classification of Diseases, Tenth Revision, codes. RESULTS: The matched cohort analysis included 4 611 378 patients, with 2 305 689 per cohort. The adjusted HR comparing NIU incidence in the COVID-19-vaccinated and unvaccinated cohort was 0.91 (95% confidence interval [CI], 0.75-1.10; P = 0.33). The SCCS analysis included 686 patients. The IRR comparing NIU risk after vaccination with risk during control intervals was 1.05 (95% CI, 0.89-1.23; P = 0.57). An increased risk was found in the subgroup aged 5 to 44 years (IRR, 1.40; 95%

Emerging from the pairwise attention in conventional Transformers, there is a growing interest in sparse attention mechanisms that align more closely with localized, contextual learning in the biological brain. Existing studies such as the Coordination method employ iterative cross-attention mechanisms with a bottleneck to enable the sparse association of inputs. However, these methods are parameter inefficient and fail in more complex relational reasoning tasks. To this end, we propose Associative Transformer (AiT) to enhance the association among sparsely attended input patches, improving parameter efficiency and performance in relational reasoning tasks. AiT leverages a learnable explicit memory, comprised of various specialized priors, with a bottleneck attention to facilitate the extraction of diverse localized features. Moreover, we propose a novel associative memory-enabled patch reconstruction with a Hopfield energy function. The extensive experiments in four image classification tasks with three different sizes of AiT demonstrate that AiT requires significantly fewer parameters and attention layers while outperforming Vision Transformers and a broad range of sparse Transformers. Additionally, AiT establishes new SOTA performance in the Sort-of-CLEVR dataset, outperforming the previous Coordination method.

BackgroundThe recombinant zoster vaccine had over 90% efficacy in preventing herpes zoster in clinical trials. However, its effectiveness outside of a clinical trial setting has not been investigated. This study aimed to assess the effectiveness of the recombinant zoster vaccine in general practice.MethodsA de-identified administrative claims database, the OptumLabs Data Warehouse, was used to conduct this retrospective cohort study to assess the effectiveness of the recombinant zoster vaccine against herpes zoster in nonimmunocompromised, vaccine age-eligible individuals enrolled in the database for ≥365 days.ResultsA total of 4 769 819 adults were included in this study, with 173 745 (3.6%) adults receiving 2 valid doses of the recombinant zoster vaccine. The incidence rate of herpes zoster was 258.8 (95% confidence interval [CI], 230.0-289.4) cases per 100 000 person-years in vaccinated persons compared with 893.1 (95% CI, 886.2-900.0) in unvaccinated persons. Recombinant zoster vaccine effectiveness was 85.5% (95% CI, 83.5-87.3%) overall, with an effectiveness of 86.8% (95% CI, 84.6-88.7%) in individuals 50 to 79 years old compared with 80.3% (95% CI, 75.1-84.3%) in individuals aged 80 and older. In patients with a history of live zoster vaccine within 5 years of study inclusion, vaccine effectiveness was 84.8% (95% CI, 75.3-90.7%).ConclusionsRecombinant zoster vaccine effectiveness against herpes zoster was high in a real-world setting. Given the low vaccine coverage and high effectiveness, a major public health effort is needed to identify and address barriers to vaccination and increase immunization rates.

BackgroundThe incidence of herpes zoster (HZ) has been on the rise for decades in the United States. Clinical trials for the recombinant zoster vaccine (RZV) demonstrated vaccine efficacy of over 90% in preventing herpes zoster. However, there is limited information on its effectiveness outside of a clinical trial setting, as well as its effectiveness against herpes zoster ophthalmicus (HZO).MethodsA de-identified electronic health records database from Kaiser Permanente Hawaii (KPH) was used to conduct this retrospective cohort study to assess the effectiveness of the recombinant zoster vaccine against HZ and HZO in immunocompetent, vaccine age-eligible individuals without a prior history of HZ, who were continuously enrolled in KPH for ≥365 days prior to becoming age-eligible for RZV between January 1, 2018, through December 31, 2019.ResultsA total of 78 356 adults were included in this study, with 11 864 (15.1%) adults receiving two valid doses of the recombinant zoster vaccine. The incidence rate of HZ was 325.6 (95% CI: 217.7 to 464.4) cases per 100 000 person-years in vaccinated persons compared to 1063.3 cases per 100 000 person-years (95% CI: 1006.0 to 1122.8) in the unvaccinated group. The incidence rate of HZO was 11.9 (95% CI: 0.7 to 52.3) cases per 100 000 person-years in the vaccinated group compared to 72.1 (95% CI: 58.0 to 88.3) in the unvaccinated group. RZV was 83.5% (95% CI: 74.9% to 89.2%) effective against HZ and 93.3% (95% CI: 48.7% to 99.1%) effective against HZO.ConclusionsRZV has demonstrated high effectiveness against both HZ and HZO outside of a clinical trial setting in the United States. Vaccine coverage is low, emphasizing the need for public health efforts to increase vaccination to reduce morbidity from HZ and HZO.

ImportanceHerpes zoster infection after COVID-19 vaccination has been reported in numerous case studies. It is not known whether these cases represent increased reporting or a true increase in risk.ObjectiveTo assess whether COVID-19 vaccination is associated with an increased risk of herpes zoster infection.Design, setting, and participantsThis cohort study used a self-controlled risk interval (SCRI) design to compare the risk of herpes zoster in a risk interval of 30 days after COVID-19 vaccination or up to the date of the second vaccine dose with a control interval remote from COVID-19 vaccination (defined as 60-90 days after the last recorded vaccination date for each individual, allowing for a 30-day washout period between control and risk intervals). A supplemental cohort analysis was used to compare the risk of herpes zoster after COVID-19 vaccination with the risk of herpes zoster after influenza vaccination among 2 historical cohorts who received an influenza vaccine in the prepandemic period (January 1, 2018, to December 31, 2019) or the early pandemic period (March 1, 2020, to November 30, 2020). Data were obtained from Optum Labs Data Warehouse, a US national deidentified claims-based database. A total of 2 039 854 individuals who received any dose of a COVID-19 vaccine with emergency use authorization (BNT162b2 [Pfizer-BioNTech], mRNA-1273 [Moderna], or Ad26.COV2.S [Johnson & Johnson]) from December 11, 2020, through June 30, 2021, were eligible for inclusion. Individuals included in the SCRI analysis were a subset of the COVID-19-vaccinated cohort who had herpes zoster during either a risk or control interval.ExposuresAny dose of a COVID-19 vaccine.Main outcomes and measuresIncident herpes zoster, defined by International Statistical Classification of Diseases and Related Health Problems, Tenth Revision codes and a prescription of a new antiviral medication or a dose increase in antiviral medication within 5 days of diagnosis.ResultsAmong 2 039 854

PurposeTo identify if noninfectious uveitis (NIU) is associated with a greater risk of Coronavirus Disease 2019 (COVID-19) infection, hospitalization, and death.DesignA retrospective cohort study from January 20, 2020 to December 31, 2020, using a national claims-based database.ParticipantsEnrollees who had continuous enrollment with both medical and pharmacy coverage for 3 years before January 20, 2020. Patients with an NIU diagnosis within 3 years of the start of the study were included in the NIU cohort. Those with infectious uveitis codes or new NIU diagnoses during the risk period were excluded.MethodsCox proportional hazard models were used to identify unadjusted hazard ratios (HRs) and adjusted HRs for all covariates for each outcome measure. Adjusted models accounted for patient demographics, health status, and immunosuppressive medication use during the risk period.Main outcome measuresRates of COVID-19 infection, COVID-19-related hospitalization, and COVID-19-related in-hospital death identified with International Classification of Disease 10th revision codes.ResultsThis study included 5 806 227 patients, of whom 29 869 (0.5%) had a diagnosis of NIU. On unadjusted analysis, patients with NIU had a higher rate of COVID-19 infection (5.7% vs. 4.5%, P < 0.001), COVID-19-related hospitalization (1.2% vs. 0.6%, P < 0.001), and COVID-19-related death (0.3% vs. 0.1%, P < 0.001). However, in adjusted models, NIU was not associated with a greater risk of COVID-19 infection (HR, 1.05; 95% confidence interval [CI], 1.00-1.10; P = 0.04), hospitalization (HR, 0.98; 95% CI, 0.88-1.09; P = 0.67), or death (HR, 0.90, 95% CI, 0.72-1.13, P = 0.37). Use of systemic corticosteroids was significantly associated with a higher risk of COVID-19 infection, hospitalization, and death.ConclusionsPatients with NIU were significantly more likely to be infected with COVID-19 and experience severe disease outcomes. However, this association was due to the demogra

The strong coupling between single quantum emitters and resonant optical micro/nanocavities is beneficial for understanding light and matter interactions. Here, we propose a plasmonic nanoantenna placed on a metal film to achieve an ultra-high electric field enhancement in the nanogap and an ultra-small optical mode volume. The strong coupling between a single quantum dot (QD) and the designed structure is investigated in detail by both numerical simulations and theoretical calculations. When a single QD is inserted into the nanogap of the silver nanoantenna, the scattering spectra show a remarkably large splitting and anticrossing behavior of the vacuum Rabi splitting, which can be achieved in the scattering spectra by optimizing the nanoantenna thickness. Our work shows another way to enhance the light/matter interaction at a single quantum emitter limit, which can be useful for many nanophotonic and quantum applications., QC 20220531

Wireless ad hoc federated learning (WAFL) is a fully decentralized collaborative machine learning framework organized by opportunistically encountered mobile nodes. Compared to conventional federated learning, WAFL performs model training by weakly synchronizing the model parameters with others, and this shows great resilience to a poisoned model injected by an attacker. In this paper, we provide our theoretical analysis of the WAFL's resilience against model poisoning attacks, by formulating the force balance between the poisoned model and the legitimate model. According to our experiments, we confirmed that the nodes directly encountered the attacker has been somehow compromised to the poisoned model but other nodes have shown great resilience. More importantly, after the attacker has left the network, all the nodes have finally found stronger model parameters combined with the poisoned model. Most of the attack-experienced cases achieved higher accuracy than the no-attack-experienced cases., Comment: 10 pages, 7 figures, to be published in IEEE International Conference on Trust, Privacy and Security in Intelligent Systems, and Applications 2022

Visual Question Answering (VQA) based on multi-modal data facilitates real-life applications such as home robots and medical diagnoses. One significant challenge is to devise a robust decentralized learning framework for various client models where centralized data collection is refrained due to confidentiality concerns. This work aims to tackle privacy-preserving VQA by decoupling a multi-modal model into representation modules and a contrastive module and leveraging inter-module gradients sharing and inter-client weight sharing. To this end, we propose Bidirectional Contrastive Split Learning (BiCSL) to train a global multi-modal model on the entire data distribution of decentralized clients. We employ the contrastive loss that enables a more efficient self-supervised learning of decentralized modules. Comprehensive experiments are conducted on the VQA-v2 dataset based on five SOTA VQA models, demonstrating the effectiveness of the proposed method. Furthermore, we inspect BiCSL's robustness against a dual-key backdoor attack on VQA. Consequently, BiCSL shows much better robustness to the multi-modal adversarial attack compared to the centralized learning method, which provides a promising approach to decentralized multi-modal learning., Comment: Accepted for AAAI 2024

Privacy-sensitive data is stored in autonomous vehicles, smart devices, or sensor nodes that can move around with making opportunistic contact with each other. Federation among such nodes was mainly discussed in the context of federated learning with a centralized mechanism in many works. However, because of multi-vendor issues, those nodes do not want to rely on a specific server operated by a third party for this purpose. In this paper, we propose a wireless ad hoc federated learning (WAFL) -- a fully distributed cooperative machine learning organized by the nodes physically nearby. WAFL can develop generalized models from Non-IID datasets stored in distributed nodes locally by exchanging and aggregating them with each other over opportunistic node-to-node contacts. In our benchmark-based evaluation with various opportunistic networks, WAFL has achieved higher accuracy of 94.8-96.3% than the self-training case of 84.7%. All our evaluation results show that WAFL can train and converge the model parameters from highly-partitioned Non-IID datasets over opportunistic networks without any centralized mechanisms., Comment: 14 pages, 8 figures, 2 tables

Model poisoning attacks on federated learning (FL) intrude in the entire system via compromising an edge model, resulting in malfunctioning of machine learning models. Such compromised models are tampered with to perform adversary-desired behaviors. In particular, we considered a semi-targeted situation where the source class is predetermined however the target class is not. The goal is to cause the global classifier to misclassify data of the source class. Though approaches such as label flipping have been adopted to inject poisoned parameters into FL, it has been shown that their performances are usually class-sensitive varying with different target classes applied. Typically, an attack can become less effective when shifting to a different target class. To overcome this challenge, we propose the Attacking Distance-aware Attack (ADA) to enhance a poisoning attack by finding the optimized target class in the feature space. Moreover, we studied a more challenging situation where an adversary had limited prior knowledge about a client's data. To tackle this problem, ADA deduces pair-wise distances between different classes in the latent feature space from shared model parameters based on the backward error analysis. We performed extensive empirical evaluations on ADA by varying the factor of attacking frequency in three different image classification tasks. As a result, ADA succeeded in increasing the attack performance by 1.8 times in the most challenging case with an attacking frequency of 0.01., Comment: Published in IJCNN 2022

Multi-source domain adaptation has been intensively studied. The distribution shift in features inherent to specific domains causes the negative transfer problem, degrading a model's generality to unseen tasks. In Federated Learning (FL), learned model parameters are shared to train a global model that leverages the underlying knowledge across client models trained on separate data domains. Nonetheless, the data confidentiality of FL hinders the effectiveness of traditional domain adaptation methods that require prior knowledge of different domain data. We propose a new federated domain generalization method called Federated Knowledge Alignment (FedKA). FedKA leverages feature distribution matching in a global workspace such that the global model can learn domain-invariant client features under the constraint of unknown client data. FedKA employs a federated voting mechanism that generates target domain pseudo-labels based on the consensus from clients to facilitate global model fine-tuning. We performed extensive experiments, including an ablation study, to evaluate the effectiveness of the proposed method in both image and text classification tasks using different model architectures. The empirical results show that FedKA achieves performance gains of 8.8% and 3.5% in Digit-Five and Office-Caltech10, respectively, and a gain of 0.7% in Amazon Review with extremely limited training data. Moreover, we studied the effectiveness of FedKA in alleviating the negative transfer of FL based on a new criterion called Group Effect. The results show that FedKA can reduce negative transfer, improving the performance gain via model aggregation by 4 times., Comment: Accepted for Asian Conference on Machine Learning (ACML 2022)

PurposeTo identify if non-infectious uveitis (NIU) is associated with a greater risk of coronavirus disease 2019 (COVID-19) infection, hospitalization, and death.DesignA retrospective cohort study from January 20, 2020 to December 31, 2020 using a national claims-based database.ParticipantsEnrollees who had continuous enrollment with both medical and pharmacy coverage for three years prior to January 20, 2020. Patients with an NIU diagnosis within three years of the start of the study were included in the NIU cohort. Those with infectious uveitis codes or new NIU diagnoses during the risk period were excluded.MethodsCox proportional hazards models were used to identify unadjusted hazard ratios as well as adjusted hazard ratios for all covariates for each outcome measure. Adjusted models accounted for patient demographics, health status, and immunosuppressive medication use during the risk period.Main outcome measuresRates of COVID-19 infection, COVID-19-related hospitalization, and COVID-19-related in-hospital death identified with International Classification of Disease 10th revision codes.ResultsThis study included 5,806,227 patients of which 29,869 (0.5%) had a diagnosis of NIU. On unadjusted analysis, patients with NIU had a higher rate of COVID-19 infection (5.7% vs. 4.2%, p<0.001), COVID-19-related hospitalization (1.2% vs. 0.6%, p<0.001) and COVID-19-related death (0.3% vs. 0.1%, p<0.001). However, in adjusted models, NIU was not associated with a greater risk of COVID-19 infection (hazard ratio [HR] = 1.05; 95% confidence interval [CI]: 1.00 - 1.10; p = 0.04), hospitalization (HR = 0.98; 95% CI: 0.88 - 1.09; p = 0.67) or death (HR = 0.90, 95% CI: 0.72 - 1.13, p = 0.37). Use of systemic corticosteroids was significantly associated with higher risk of COVID-19 infection, hospitalization, and death.ConclusionsPatients with NIU were significantly more likely to be infected with COVID-19 and experience severe disease outcomes. However, this association w

Compared to natural materials, artificial subwavelength structures can enhance chiroptical effects in a stronger way, and the requirement of low material loss and wideband operation is desired in many situations. Here, we propose an all-dielectric chiral metasurface as a periodic array of centrosymmetric staggered silicon cuboid pairs to achieve strong circular dichroism in a wide band. As a demonstration, the designed chiral metasurface may strongly reflect the chosen circularly polarized light with the spin preserved in the operating wavelength range of 1.51 ti 1.60 mu m while highly transmit (with an efficiency greater than 95%) the opposite circularly polarized light with the spin flipped. Then, two application cases are given for the designed chiral metasurface. A flat chiral meta-lens is constructed to produce wideband focusing in the transmission/reflection side while the disturbing from the opposite circular polarization is well blocked by high reflection/transmission. A chiral Fabry-Perot cavity is also constructed, which has an extremely high quality factor (2.1E4). The proposed method provides an efficient way to produce strong chiroptical effects and has a promising potential for various applications such as signal processing, sensing, radiation and detection., QC 20220308

PurposeTo examine the effectiveness of the recombinant zoster vaccine (RZV) for preventing herpes zoster ophthalmicus (HZO) in the general United States population.DesignRetrospective, observational cohort study.ParticipantsIndividuals enrolled in the OptumLabs Data Warehouse (OLDW; OptumLabs, Cambridge, MA) who were age eligible for herpes zoster (HZ) vaccination (≥50 years of age) from 2018 through 2019. The OLDW is a longitudinal, de-identified administrative claims and electronic health record database of patients in the United States with commercial insurance, Medicare Part D, or Medicare Advantage METHODS: Patients were required to have 365 days or more of continuous enrollment to be eligible. Those with a diagnosis code of HZ or an immunocompromising condition within 1 year before study inclusion were excluded. Vaccination with the RZV was ascertained by Current Procedural Terminology codes, and HZO was ascertained by International Classification of Diseases, Tenth Revision, codes. Cox proportional hazards regression models were used to estimate the hazard ratio of HZO associated with RZV, and inverse-probability weighting was used to control for confounding. Vaccine effectiveness was calculated from hazard ratios.Main outcome measuresIncidence of HZO in vaccinated versus unvaccinated person-times and vaccine effectiveness were assessed.ResultsFrom January 1, 2018, through December 31, 2019, a total of 4 842 579 individuals were included in this study. One hundred seventy-seven thousand two hundred eighty-nine (3.7%) received 2 valid doses of RZV. The incidence rate of HZO was 25.5 cases (95% confidence interval [CI], 17.4-35.8 cases) per 100 000 person-years in the vaccinated group compared with 76.7 cases (95% CI, 74.7-78.7 cases) in the unvaccinated group. The overall adjusted effectiveness of RZV against HZO was 89.1% (95% CI, 82.9%-93.0%).ConclusionsThe effectiveness of RZV against HZO in individuals 50 years of age and older is high in a clinical settin

Federated learning (FL) has been facilitating privacy-preserving deep learning in many walks of life such as medical image classification, network intrusion detection, and so forth. Whereas it necessitates a central parameter server for model aggregation, which brings about delayed model communication and vulnerability to adversarial attacks. A fully decentralized architecture like Swarm Learning allows peer-to-peer communication among distributed nodes, without the central server. One of the most challenging issues in decentralized deep learning is that data owned by each node are usually non-independent and identically distributed (non-IID), causing time-consuming convergence of model training. To this end, we propose a decentralized learning model called Homogeneous Learning (HL) for tackling non-IID data with a self-attention mechanism. In HL, training performs on each round's selected node, and the trained model of a node is sent to the next selected node at the end of each round. Notably, for the selection, the self-attention mechanism leverages reinforcement learning to observe a node's inner state and its surrounding environment's state, and find out which node should be selected to optimize the training. We evaluate our method with various scenarios for an image classification task. The result suggests that HL can produce a better performance compared with standalone learning and greatly reduce both the total training rounds by 50.8% and the communication cost by 74.6% compared with random policy-based decentralized learning for training on non-IID data.

With increasingly more sophisticated phishing campaigns in recent years, phishing emails lure people using more legitimate-looking personal contexts. To tackle this problem, instead of traditional heuristics-based algorithms, more adaptive detection systems such as natural language processing (NLP)-powered approaches are essential to understanding phishing text representations. Nevertheless, concerns surrounding the collection of phishing data that might cover confidential information hinder the effectiveness of model learning. We propose a decentralized phishing email detection framework called Federated Phish Bowl (FedPB) which facilitates collaborative phishing detection with privacy. In particular, we devise a knowledge-sharing mechanism with federated learning (FL). Using long short-term memory (LSTM) for phishing detection, the framework adapts by sharing a global word embedding matrix across the clients, with each client running its local model with Non-IID data. We collected the most recent phishing samples to study the effectiveness of the proposed method using different client numbers and data distributions. The results show that FedPB can attain a competitive performance with a centralized phishing detector, with generality to various cases of FL retaining a prediction accuracy of 83%., Comment: To be published in 2022 IEEE International Conference on Systems, Man, and Cybernetics (SMC)

The rapidly increasing number of smart devices on the Internet necessitates an efficient inspection system for safeguarding our networks from suspicious activities such as Address Resolution Protocol (ARP) probes. In this research, we analyze sequence data of ARP traffic on LAN based on the numerical count and degree of its packets. Moreover, a dynamic threshold is employed to detect underlying suspicious activities, which are further converted into feature vectors to train an unsupervised autoencoder neural network. Then, we leverage K-means clustering to separate the extracted latent features of suspicious activities from the autoencoder into various patterns. Besides, to evaluate the performance, we collect and adopt a real-world network traffic dataset from five different LANs. At last, we successfully detect suspicious ARP patterns varying in scale, lifespan, and regularity on the LANs., Comment: 5 pages, 7 figures, submitted to 2022 IEEE Consumer Communications & Networking Conference

Wider coverage and a better solution to a latency reduction in 5G necessitate its combination with multi-access edge computing (MEC) technology. Decentralized deep learning (DDL) such as federated learning and swarm learning as a promising solution to privacy-preserving data processing for millions of smart edge devices, leverages distributed computing of multi-layer neural networks within the networking of local clients, whereas, without disclosing the original local training data. Notably, in industries such as finance and healthcare where sensitive data of transactions and personal medical records is cautiously maintained, DDL can facilitate the collaboration among these institutes to improve the performance of trained models while protecting the data privacy of participating clients. In this survey paper, we demonstrate the technical fundamentals of DDL that benefit many walks of society through decentralized learning. Furthermore, we offer a comprehensive overview of the current state-of-the-art in the field by outlining the challenges of DDL and the most relevant solutions from novel perspectives of communication efficiency and trustworthiness., Comment: 11 pages, 8 figures, to appear in IEEE Transactions on Artificial Intelligence

An attack on deep learning systems where intelligent machines collaborate to solve problems could cause a node in the network to make a mistake on a critical judgment. At the same time, the security and privacy concerns of AI have galvanized the attention of experts from multiple disciplines. In this research, we successfully mounted adversarial attacks on a federated learning (FL) environment using three different datasets. The attacks leveraged generative adversarial networks (GANs) to affect the learning process and strive to reconstruct the private data of users by learning hidden features from shared local model parameters. The attack was target-oriented drawing data with distinct class distribution from the CIFAR- 10, MNIST, and Fashion-MNIST respectively. Moreover, by measuring the Euclidean distance between the real data and the reconstructed adversarial samples, we evaluated the performance of the adversary in the learning processes in various scenarios. At last, we successfully reconstructed the real data of the victim from the shared global model parameters with all the applied datasets., Comment: 7 pages, 11 figures, to be published in proceedings of IEEE International Conference on Systems, Man, and Cybernetics (SMC) 2021

This project tried to create a new pedestrian bridge on edge of Vedado, which also act as new public space of the urban area. And this project would offer a new place of Cuba to have fun with water., Havana Studio, Architecture, Urbanism and Building Sciences | Complex Projects

This project tried to create a new pedestrian bridge on edge of Vedado, which also act as new public space of the urban area. And this project would offer a new place of Cuba to have fun with water., Havana Studio, Architecture, Urbanism and Building Sciences | Complex Projects