Federated Learning Model for Auto Insurance Rate Setting Based on Tweedie Distribution

In the assessment of car insurance claims,the claim rate for car insurance presents a highly skewed probability distribution,which is typically modeled using Tweedie distribution.The traditional approach to obtaining the Tweedie regression model involves training on a centralized dataset,when the data is provided by multiple parties,training a privacy-preserving Tweedie regression model without exchanging raw data becomes a challenge.To address this issue,this study introduces a novel vertical federated learning-based Tweedie regression algorithm for multi-party auto insurance rate setting in data silos.The algorithm can keep sensitive data locally and uses privacy-preserving techniques to achieve intersection operations between the two parties holding the data.After determining which entities are shared,the participants train the model locally using the shared entity data to obtain the local generalized linear model intermediate parameters.The homomorphic encryption algorithms are introduced to interact with and update the model intermediate parameters to collaboratively complete the joint training of the car insurance rate-setting model.Performance tests on two publicly available datasets show that the proposed federated Tweedie regression algorithm can effectively generate Tweedie regression models that leverage the value of data fromboth partieswithout exchanging data.The assessment results of the scheme approach those of the Tweedie regressionmodel learned fromcentralized data,and outperformthe Tweedie regressionmodel learned independently by a single party.

Effects of mesenchymal stem cell on dopaminergic neurons,motor and memory functions in animal models of Parkinson's disease:a systematic review and meta-analysis

Parkinson’s disease is chara cterized by the loss of dopaminergic neurons in the substantia nigra pars com pacta,and although restoring striatal dopamine levels may improve symptoms,no treatment can cure or reve rse the disease itself.Stem cell therapy has a regenerative effect and is being actively studied as a candidate for the treatment of Parkinson’s disease.Mesenchymal stem cells are considered a promising option due to fewer ethical concerns,a lower risk of immune rejection,and a lower risk of teratogenicity.We performed a meta-analysis to evaluate the therapeutic effects of mesenchymal stem cells and their derivatives on motor function,memory,and preservation of dopamine rgic neurons in a Parkinson’s disease animal model.We searched bibliographic databases(PubMed/MEDLINE,Embase,CENTRAL,Scopus,and Web of Science)to identify articles and included only pee r-reviewed in vivo interve ntional animal studies published in any language through J une 28,2023.The study utilized the random-effect model to estimate the 95%confidence intervals(CI)of the standard mean differences(SMD)between the treatment and control groups.We use the systematic review center for laboratory animal expe rimentation’s risk of bias tool and the collaborative approach to meta-analysis and review of animal studies checklist for study quality assessment.A total of 33studies with data from 840 Parkinson’s disease model animals were included in the meta-analysis.Treatment with mesenchymal stem cells significantly improved motor function as assessed by the amphetamine-induced rotational test.Among the stem cell types,the bone marrow MSCs with neurotrophic factor group showed la rgest effect size(SMD[95%CI]=-6.21[-9.50 to-2.93],P=0.0001,I^(2)=0.0%).The stem cell treatment group had significantly more tyrosine hydroxylase positive dopamine rgic neurons in the striatum([95%CI]=1.04[0.59 to 1.49],P=0.0001,I^(2)=65.1%)and substantia nigra(SMD[95%CI]=1.38[0.89 to 1.87],P=0.0001,I^(2)=75.3%),indicating a protective effect on dopaminergic neurons.Subgroup analysis of the amphetamine-induced rotation test showed a significant reduction only in the intracranial-striatum route(SMD[95%CI]=-2.59[-3.25 to-1.94],P=0.0001,I^(2)=74.4%).The memory test showed significant improvement only in the intravenous route(SMD[95%CI]=4.80[1.84 to 7.76],P=0.027,I^(2)=79.6%).Mesenchymal stem cells have been shown to positively impact motor function and memory function and protect dopaminergic neurons in preclinical models of Parkinson’s disease.Further research is required to determine the optimal stem cell types,modifications,transplanted cell numbe rs,and delivery methods for these protocols.

Trusted Encrypted Traffic Intrusion Detection Method Based on Federated Learning and Autoencoder

With the rapid development of the Internet,network security and data privacy are increasingly valued.Although classical Network Intrusion Detection System(NIDS)based on Deep Learning(DL)models can provide good detection accuracy,but collecting samples for centralized training brings the huge risk of data privacy leakage.Furthermore,the training of supervised deep learning models requires a large number of labeled samples,which is usually cumbersome.The“black-box”problem also makes the DL models of NIDS untrustworthy.In this paper,we propose a trusted Federated Learning(FL)Traffic IDS method called FL-TIDS to address the above-mentioned problems.In FL-TIDS,we design an unsupervised intrusion detection model based on autoencoders that alleviates the reliance on marked samples.At the same time,we use FL for model training to protect data privacy.In addition,we design an improved SHAP interpretable method based on chi-square test to perform interpretable analysis of the trained model.We conducted several experiments to evaluate the proposed FL-TIDS.We first determine experimentally the structure and the number of neurons of the unsupervised AE model.Secondly,we evaluated the proposed method using the UNSW-NB15 and CICIDS2017 datasets.The exper-imental results show that the unsupervised AE model has better performance than the other 7 intrusion detection models in terms of precision,recall and f1-score.Then,federated learning is used to train the intrusion detection model.The experimental results indicate that the model is more accurate than the local learning model.Finally,we use an improved SHAP explainability method based on Chi-square test to analyze the explainability.The analysis results show that the identification characteristics of the model are consistent with the attack characteristics,and the model is reliable.

CCM-FL:Covert communication mechanisms for federated learning in crowd sensing IoT

The past decades have witnessed a wide application of federated learning in crowd sensing,to handle the numerous data collected by the sensors and provide the users with precise and customized services.Meanwhile,how to protect the private information of users in federated learning has become an important research topic.Compared with the differential privacy(DP)technique and secure multiparty computation(SMC)strategy,the covert communication mechanism in federated learning is more efficient and energy-saving in training the ma-chine learning models.In this paper,we study the covert communication problem for federated learning in crowd sensing Internet-of-Things networks.Different from the previous works about covert communication in federated learning,most of which are considered in a centralized framework and experimental-based,we firstly proposes a centralized covert communication mechanism for federated learning among n learning agents,the time complexity of which is O(log n),approximating to the optimal solution.Secondly,for the federated learning without parameter server,which is a harder case,we show that solving such a problem is NP-hard and prove the existence of a distributed covert communication mechanism with O(log logΔlog n)times,approximating to the optimal solution.Δis the maximum distance between any pair of learning agents.Theoretical analysis and nu-merical simulations are presented to show the performance of our covert communication mechanisms.We hope that our covert communication work can shed some light on how to protect the privacy of federated learning in crowd sensing from the view of communications.

A Model of Grand Unified Theory: Suggested Solution for CP-Violation Using Ideas of Phase Paths

This study demonstrates that beyond standard model (BSM) cosmic fundamental interactions—weak, strong, and electromagnetic forces—can be unified through a common basis of representation. This unification allows for the derivation of the fine structure constant with running points of α(t) ≈ 1/(136.9038) at high energy scales, based on electroweak interactions. Through the application of the Ising model, the running point of the elementary charge e at high energy scales is determined, and Coulomb’s law is actually derived from the Yukawa potential. Theoretically, based on S. Weinberg’s electroweak interaction theory, this study unifies the strong and electromagnetic forces by representing them with rYuka, and further advances the reconstruction of the SU(3)C×SU(1)L×U(1)EMframework on the basis of electroweak interaction concepts. In fact, the cosmic fundamental forces can interchange at the mass gap, defined as the Yukawa turning phase at rYuka ≃1.9404 fm, with the SU(3)Diag structural constant fijk on glueballs calculated, estimating a spectrum mass gap of ∆0 > 0.

Evaluating the role of large language models in inflammatory bowel disease patient information

This letter evaluates the article by Gravina et al on ChatGPT’s potential in providing medical information for inflammatory bowel disease patients.While promising,it highlights the need for advanced techniques like reasoning+action and retrieval-augmented generation to improve accuracy and reliability.Emphasizing that simple question and answer testing is insufficient,it calls for more nuanced evaluation methods to truly gauge large language models’capabilities in clinical applications.

Solving the Olbers’s Paradox, Explaining the “Red-Shift”, and Challenging the Relativities by “Sun Matters Theory” and “Sun Model of Universe”, an Evolution of the Einstein’s Static Universe Model

Olbers’s paradox, known as the dark night paradox, is an argument in astrophysics that the darkness of the night sky conflicts with the assumption of an infinite and eternal static universe. Big-Bang theory was used to partially explain this paradox, while introducing new problems. Hereby, we propose a better theory, named Sun Matters Theory, to explain this paradox. Moreover, this unique theory supports and extended the Einstein’s static universe model proposed by Albert Einstein in 1917. Further, we proposed our new universe model, “Sun Model of Universe”. Based on the new model and novel theory, we generated innovative field equation by upgrading Einstein’s Field Equation through adding back the cosmological constant, introducing a new variable and modifying the gravitationally-related concepts. According to the Sun Model of Universe, the dark matter and dark energy comprise the so-called “Sun Matters”. The observed phenomenon like the red shift is explained as due to the interaction of ordinary light with Sun Matters leading to its energy and frequency decrease. In Sun Model, our big universe consists of many universes with ordinary matter at the core mixed and surrounded with the Sun Matters. In those universes, the laws of physics may be completely or partially different from that of our ordinary universe with parallel civilizations. The darkness of night can be easily explained as resulting from the interaction of light with the Sun Matters leading to the sharp decrease in the light intensity. Sun Matters also scatter the light from a star, which makes it shining as observed by Hubble. Further, there is a kind of Sun Matters named “Sun Waters”, surrounding every starts. When lights pass by the sun, the Sun Waters deflect the lights to bend the light path. According to the Sun Model, it is the light bent not the space bent that was proposed in the theory of relativities.

Hubble Tension Explanation from This Cosmological Model AΛΩ (Slow Bang Model, SB)

In this article we present a model of Hubble-Lemaître law using the notions of a transmitter (galaxy) and a receiver (MW) coupled to a model of the universe (Slow Bang Model, SB), based on a quantum approach of the evolution of space-time as well as an equation of state that retains all the infinitesimal terms. We find an explanation of the Hubble tension H0. Indeed, we have seen that this constant depends on the transceiver pair which can vary from the lowest observable value, from photons of the CMB (theoretical [km/s/Mpc]) to increasingly higher values depending on the earlier origin of the formation of the observed galaxy or cluster (ETG ~0.3 [Gy], ~74 [km/s/Mpc]). We have produced a theoretical table of the values of the constant according to the possible pairs of transmitter/receiver in the case where these galaxies follow the Hubble flow without large disturbance. The calculated theoretical values of the constant are in the order of magnitude of all values mentioned in past studies. Subsequently, we applied the models to 9 galaxies and COMA cluster and found that the models predict acceptable values of their distances and Hubble constant since these galaxies mainly follow the Hubble flow rather than the effects of a galaxy cluster or a group of clusters. In conclusion, we affirm that this Hubble tension does not really exist and it is rather the understanding of the meaning of this constant that is questioned.

Analysis of Porter’s Five Forces Competitive Model in China’s Property Rights Trading Market

In recent years,China’s property rights trading market has been extremely competitive,but there are also new contradictions and challenges.This paper aims to analyze the external and internal competitive market of the property rights market through Porter’s five forces competition model,then find out the problems and defects in the development,thereby promoting the better development and progress of China’s property rights trading market.

Real-Time Prediction Algorithm for Intelligent Edge Networks with Federated Learning-Based Modeling

Intelligent healthcare networks represent a significant component in digital applications,where the requirements hold within quality-of-service(QoS)reliability and safeguarding privacy.This paper addresses these requirements through the integration of enabler paradigms,including federated learning(FL),cloud/edge computing,softwaredefined/virtualized networking infrastructure,and converged prediction algorithms.The study focuses on achieving reliability and efficiency in real-time prediction models,which depend on the interaction flows and network topology.In response to these challenges,we introduce a modified version of federated logistic regression(FLR)that takes into account convergence latencies and the accuracy of the final FL model within healthcare networks.To establish the FLR framework for mission-critical healthcare applications,we provide a comprehensive workflow in this paper,introducing framework setup,iterative round communications,and model evaluation/deployment.Our optimization process delves into the formulation of loss functions and gradients within the domain of federated optimization,which concludes with the generation of service experience batches for model deployment.To assess the practicality of our approach,we conducted experiments using a hypertension prediction model with data sourced from the 2019 annual dataset(Version 2.0.1)of the Korea Medical Panel Survey.Performance metrics,including end-to-end execution delays,model drop/delivery ratios,and final model accuracies,are captured and compared between the proposed FLR framework and other baseline schemes.Our study offers an FLR framework setup for the enhancement of real-time prediction modeling within intelligent healthcare networks,addressing the critical demands of QoS reliability and privacy preservation.

高压管汇材料疲劳性能测试及P-S-N模型曲线的拟合

高压管汇作为压裂设备中的主要易损件之一,其失效危害较大。它的失效原因主要是疲劳、冲蚀、腐蚀或者材料缺陷引起的刺漏和爆裂,其中尤以疲劳失效最不可预估。目前,对于高压管汇材料的疲劳性能研究不够深入,为解决高压管汇材料疲劳寿命的准确描述问题,以某国产高压管汇材料为例,进行了一系列疲劳试验,并基于试验数据,采用多种分布模型和不同S-N模型进行拟合分析,得出综合评价拟合能力最强的P-S-N模型。结果表明,该材料在中长疲劳寿命区,Weibull三参数模型在7级应力水平下综合评价能力最好;在存活率分别为50%、90%、99%、99.9%时,指数S-N模型的拟合系数均大于0.98,拟合能力最好。得出的P-S-N模型曲线可以为高压管汇的疲劳寿命以及安全设计提供依据。

多IRST自适应协同定位技术研究

为解决传统单红外搜索与跟踪(IRST)系统观测度不足,无法提供目标全维态势信息的问题,提出了基于几何精度因子(GDOP)的多IRST自适应协同定位技术。其核心思想为对多IRST进行双机组合配对,依据配对组合的GDOP值构造面向定位精度的信息类函数,进而确定联邦滤波信息分配系数,最后基于自适应联邦滤波算法解决双IRST配对组合局部估计的全局信息融合问题,得到目标的精确位置信息。通过仿真结果证明,所提算法定位精度高、可靠性强,同时,可保证全局定位结果的精确性及稳定性。

深水采油树Spring-Seal密封圈接触模型构建与实验测试

为提高深水采油树在高压、高温、低温等工况下长时间工作的稳定性,对采油树密封圈进行重新设计,提出Spring-Seal密封圈(以下简称S密封圈);基于赫兹接触理论、填料密封接触理论,重新构建S密封圈接触力学模型;建立有限元仿真分析模型,分析不同条件下S密封圈的密封性能,同时验证新建接触模型的合理性;利用实验测试S密封圈的实际密封效果。研究结果表明:S密封圈在高压、高温、低温等工况下能够长时间稳定工作且密封性能良好。研究结果可为S密封圈的接触研究及工程运用提供理论参考。

D-S理论和Markov链组合的桥梁性能退化预测研究

为准确预测桥梁性能退化,考虑到数据随机性和微小扰动发生状态跳跃,提出了一种D-S(Dempster-Shafer)证据理论和Markov链组合的桥梁性能退化组合预测模型和性能退化率的概念.该模型基于指数平滑(exponential smoothing,ES)方法获得新的预测数据序列,并利用Markov链和D-S理论不断进行优化,从而实现桥梁性能退化的组合预测.实际工程的应用结果表明:性能退化率可以直观地表征在梁性能退化的速度.其次,该模型的平均相对误差为1.54%,较于回归、灰色和模糊加权Markov链模型,精度分别提高了1.11%,0.88%和2.8%,而后验差比值为0.242,小于0.35;模型的标准差为9.021,相比其他模型分别减小了3.978,3.405和7.500,而变异系数为0.109,均小于其他模型,验证了组合预测模型在精度和稳定性方面的优越性,可为在役桥梁结构性能退化预测与维护提供理论基础.

基于量子衍生涡流算法和T⁃S模糊推理模型的储层岩性识别

鉴于梯度下降法易陷入局部极值、普通群智能优化算法易早熟收敛,提出一种基于量子衍生涡流算法(Quantum Vortex Search Algorithm,QVSA)和T⁃S模糊推理模型的岩性识别方法,QVSA具有操作简单、收敛速度快、寻优能力强等优点,有助于T⁃S模糊推理模型获得最优参数配置,从而实现储层岩性的准确识别。首先利用具有全局搜索能力的QVSA优化T⁃S模糊推理模型的各种参数;然后利用主成分分析方法降低获取的地震属性维度;再利用优化的T⁃S模糊推理模型识别储层岩性。实验结果表明,利用反映储层特征的8个地震属性识别储层岩性时,所提方法的识别正确率达到92%,比普通BP网络方法高5.1%,同时查准率、查全率、F1分数等指标也较BP网络方法提升明显。

基于S变换模能量分析的直流电缆故障测距

针对现有高压直流电缆故障定位方法故障定位精度低、定位模型泛化能力差等问题,文中在考虑电缆故障后故障点产生的暂态信号中含有丰富的故障信息的特征,提出一种时频域能量分析的故障测距方法。该方法利用多分类支持向量机(SVM)的特性构建直流电缆故障定位模型,利用直流电缆相模解耦矩阵对电缆电气量解耦,并采用S变换进行时频域变换,将得到的模量构造模能量,将其作为样本对SVM模型训练。实验结果表明,所提方法对于不同故障类型在不同过渡电阻及不同位置故障下均可进行准确测距。

基于S-O-R模型的视觉营销对消费者购买意愿的影响研究——以红色文创网店为例

为提高红色文创网店的点击转化率,文章从消费者视角出发,研究影响消费者购买意愿的视觉因素。基于S-O-R模型理论,构建红色文创网店视觉营销对消费者购买意愿因素的理论模型。利用S-O-R模型分析了红色文创网店视觉营销中的商品形象、界面形象以及品牌视觉形象对顾客感知价值及顾客购买意向的影响。通过设计量表和调查问卷进行实证研究,并结合结构方程模型加以论证,得出品牌视觉形象和界面形象是影响消费者购买意愿的关键因素,为设计师和红色文创品牌进行网店设计提供了一定的理论参考和建议。

S-T-S混合式循环互助教学模式用于少学时“化工原理”课程

“化工原理”课程是化工及相关专业的专业基础课程,相关专业的学时少,课程难,需重点解决的问题是“三多”“三不愿”“三不明”。针对痛点问题,提出了线上线下混合式循环互助的创新设计思路,包括:教师帮助学生自制课前、课中、课后习题,学生帮助教师反馈真实学习情况,学生帮助学生组团互助学习。问卷调查和评教结果显示,在一定程度上解决了“三多”“三不愿”的问题;通过建立“助教团队”反馈不同层次学生的学习情况,较好地解决了“三不明”的问题。学生的成绩与教改前相比有明显提高。

U-G-S模式下现代大学附属学校建设的路径思考

U-G-S(大学-政府-学校)三方联动是上海师范大学近年来建设现代大学附属学校的主要模式,上海师范大学、各区政府与中小学校高效协同合作,发挥各方优势,主动服务上海市政府建设“五个新城”的城市战略,办老百姓家门口的好学校。大学附属学校分为挂牌学校和新建学校两种类型,研究从内在机制、实践路径和建设成效三个方面对两类学校的变革发展进行深度剖析,以期为大学附属学校建设提供宝贵经验。

Outlook of PINK1/Parkin signaling in molecular etiology of Parkinson's disease,with insights into Pink1 knockout models

Parkinson’s disease(PD)relates to defective mitochondrial quality control in the dopaminergic motor network.Genetic studies have revealed that PINK1 and Parkin mutations are indicative of a heightened propensity to PD onset,pinpointing mitophagy and inflammation as the culprit pathways involved in neuronal loss in the substantia nigra(SNpc).In a reciprocal manner,LRRK2 functions in the regulation of basal flux and inflammatory responses responsible for PINK1/Parkin-dependent mitophagy activation.Pharmacological intervention in these diseasemodifying pathways may facilitate the development of novel PD therapeutics,despite the current lack of an established drug evaluation model.As such,we reviewed the feasibility of employing the versatile global Pink1knockout(KO)rat model as a self-sufficient,spontaneous PD model for investigating both disease etiology and drug pharmacology.These rats retain clinical features encompassing basal mitophagic flux changes with PD progression.We demonstrate the versatility of this PD rat model based on the incorporation of additional experimental insults to recapitulate the proinflammatory responses observed in PD patients.