To solve the problem of increased computation and communication costs caused by using homomorphic encryption(HE) to protect all gradients in traditional cryptographic aggregation(cryptoaggregation) schemes,a fast cryp...To solve the problem of increased computation and communication costs caused by using homomorphic encryption(HE) to protect all gradients in traditional cryptographic aggregation(cryptoaggregation) schemes,a fast crypto-aggregation scheme called RandomCrypt was proposed.RandomCrypt performed clipping and quantization to fix the range of gradient values and then added two types of noise on the gradient for encryption and differential privacy(DP) protection.It conducted HE on noise keys to revise the precision loss caused by DP protection.RandomCrypt was implemented based on a FATE framework,and a hacking simulation experiment was conducted.The results show that the proposed scheme can effectively hinder inference attacks while ensuring training accuracy.It only requires 45%~51% communication cost and 5%~23% computation cost compared with traditional schemes.展开更多
In this paper, a new method of time reversal for defect diagnosis of concealed structure has been proposed based on the detecting technique of structure acoustic wave and the theory of time reversal. The time reversal...In this paper, a new method of time reversal for defect diagnosis of concealed structure has been proposed based on the detecting technique of structure acoustic wave and the theory of time reversal. The time reversal recurrence formula for detecting the acoustic wave speed constitution of concealed structures with bilevel asynchronous test has been established. The wave speed constitution can be reconstructed in 2 D graticule form by means of this method. The result of model test shows the method is valid.展开更多
In this paper,we prove existence results of soutions for the nonlinear implicit complementarity problems NICP(T,S,K) where K is a closed weakly locally compact convex cone in a reflexive Banach space E,T is a nonlinea...In this paper,we prove existence results of soutions for the nonlinear implicit complementarity problems NICP(T,S,K) where K is a closed weakly locally compact convex cone in a reflexive Banach space E,T is a nonlinear operator from K into E* (i. e.,the dual space of E) and S is a nonlinear operator from K into E. Our results are the essential improvements and extension of the results obtained previously by several authors including Thera,Ding,and Zeng.展开更多
Implicit feedback, which indirectly reflects opinion through user behaviors, has gained increasing attention in recommender system communities due to its accessibility and richness in real-world applications. A major ...Implicit feedback, which indirectly reflects opinion through user behaviors, has gained increasing attention in recommender system communities due to its accessibility and richness in real-world applications. A major way of exploiting implicit feedback is to treat the data as an indication of positive and negative preferences associated with vastly varying confidence levels. Such algorithms assume that the numerical value of implicit feedback, such as time of watching, indicates confidence, rather than degree of preference, and a larger value indicates a higher confidence, although this works only when just one type of implicit feedback is available. However, in real-world applications, there are usually various types of implicit feedback, which can be referred to as heterogeneous implicit feedback. Existing methods cannot efficiently infer confidence levels from heterogeneous implicit feedback. In this paper, we propose a novel confidence estimation approach to infer the confidence level of user preference based on heterogeneous implicit feedback. Then we apply the inferred confidence to both point-wise and pair-wise matrix factorization models, and propose a more generic strategy to select effective training samples for pair-wise methods. Experiments on real-world e-commerce datasets from Tmall.com show that our methods outperform the state-of-the-art approaches, consid- ering several commonly used ranking-oriented evaluation criteria.展开更多
Because antipredator behaviours are costly, the threat-sensitive predator avoidance hypothesis predicts that individual animals should express predator-avoidance behaviour proportionally to the perceived threat posed ...Because antipredator behaviours are costly, the threat-sensitive predator avoidance hypothesis predicts that individual animals should express predator-avoidance behaviour proportionally to the perceived threat posed by the predator. Here, we experimentally tested this hypothesis by providing wild passerine birds supplemental food （on a raised feeding platform） at either 1 or 4 m from the edge of forest cover （potential refuge）, in either the presence or absence of a nearby simulated predation threat （a sharp-shinned hawk Accipiter striatus model）. Compared with the control treatment, we observed proportionally fewer bird visits to the food patch, and the birds took longer to re-emerge from forest refuge and return to feed at the food patch, after the hawk presentation than before it. The observed threat-sensitive latency-to-return response was stronger when the food patch was further away from the nearest refuge. Overall, our results are consistent with the predictions of the threat-sensitive predator avoidance hypothesis in that wild passerine birds （primarily black-capped chickadees Poecile atricapillus） exhibited more intense antipre- dator behavioural responses with increasing level of apparent threat. The birds were thus sensitive to their local perceived threat of predation and traded-off safety from predation （by refuging） and foraging gains in open habitat in a graded, threat-sensitive manner [Current Zoology 60 （6）： 719-728, 2014].展开更多
This paper aims to investigate the effects of recasts and working memory on the acquisition of Korean morphological causatives by advanced Chinese learners of Korean. Participants were randomly assigned to two groups...This paper aims to investigate the effects of recasts and working memory on the acquisition of Korean morphological causatives by advanced Chinese learners of Korean. Participants were randomly assigned to two groups： A experimental group and a control group. The experimental group received intensive recasts during task-based interaction, whereas the control group did not. The effects of recasts were measured by two types of tests： An elicited imitation test, as a measure of implicit knowledge, and an untimed grammaticality judgment test, as a measure of explicit knowledge. The findings are as follows. First, from the pretest to the delayed posttest, recasts facilitated the learners＇ acquisition of causative construction by developing both their implicit and explicit knowledge. Second, the results of delayed posttests showed that recasts were more effective in aiding the development of implicit knowledge than explicit knowledge. Third, working memory was found to be a significant covariate in the facilitative effects of recasts on explicit knowledge; that is, it mediated the development of explicit knowledge via recasts as a significant individual difference factor. The effects of recasts were maintained even when the impact of working memory was controlled. （189 words）展开更多
基金Beijing Natural Science Foundation (L233005)National Key Research and Development Program of China (2023YFB3308200)。
文摘To solve the problem of increased computation and communication costs caused by using homomorphic encryption(HE) to protect all gradients in traditional cryptographic aggregation(cryptoaggregation) schemes,a fast crypto-aggregation scheme called RandomCrypt was proposed.RandomCrypt performed clipping and quantization to fix the range of gradient values and then added two types of noise on the gradient for encryption and differential privacy(DP) protection.It conducted HE on noise keys to revise the precision loss caused by DP protection.RandomCrypt was implemented based on a FATE framework,and a hacking simulation experiment was conducted.The results show that the proposed scheme can effectively hinder inference attacks while ensuring training accuracy.It only requires 45%~51% communication cost and 5%~23% computation cost compared with traditional schemes.
文摘In this paper, a new method of time reversal for defect diagnosis of concealed structure has been proposed based on the detecting technique of structure acoustic wave and the theory of time reversal. The time reversal recurrence formula for detecting the acoustic wave speed constitution of concealed structures with bilevel asynchronous test has been established. The wave speed constitution can be reconstructed in 2 D graticule form by means of this method. The result of model test shows the method is valid.
文摘In this paper,we prove existence results of soutions for the nonlinear implicit complementarity problems NICP(T,S,K) where K is a closed weakly locally compact convex cone in a reflexive Banach space E,T is a nonlinear operator from K into E* (i. e.,the dual space of E) and S is a nonlinear operator from K into E. Our results are the essential improvements and extension of the results obtained previously by several authors including Thera,Ding,and Zeng.
基金supported by the National Basic Research Program(973) of China(No.2015CB352400)the National Key Research and Development Program of China(No.2016YFB1200203-03)
文摘Implicit feedback, which indirectly reflects opinion through user behaviors, has gained increasing attention in recommender system communities due to its accessibility and richness in real-world applications. A major way of exploiting implicit feedback is to treat the data as an indication of positive and negative preferences associated with vastly varying confidence levels. Such algorithms assume that the numerical value of implicit feedback, such as time of watching, indicates confidence, rather than degree of preference, and a larger value indicates a higher confidence, although this works only when just one type of implicit feedback is available. However, in real-world applications, there are usually various types of implicit feedback, which can be referred to as heterogeneous implicit feedback. Existing methods cannot efficiently infer confidence levels from heterogeneous implicit feedback. In this paper, we propose a novel confidence estimation approach to infer the confidence level of user preference based on heterogeneous implicit feedback. Then we apply the inferred confidence to both point-wise and pair-wise matrix factorization models, and propose a more generic strategy to select effective training samples for pair-wise methods. Experiments on real-world e-commerce datasets from Tmall.com show that our methods outperform the state-of-the-art approaches, consid- ering several commonly used ranking-oriented evaluation criteria.
文摘Because antipredator behaviours are costly, the threat-sensitive predator avoidance hypothesis predicts that individual animals should express predator-avoidance behaviour proportionally to the perceived threat posed by the predator. Here, we experimentally tested this hypothesis by providing wild passerine birds supplemental food （on a raised feeding platform） at either 1 or 4 m from the edge of forest cover （potential refuge）, in either the presence or absence of a nearby simulated predation threat （a sharp-shinned hawk Accipiter striatus model）. Compared with the control treatment, we observed proportionally fewer bird visits to the food patch, and the birds took longer to re-emerge from forest refuge and return to feed at the food patch, after the hawk presentation than before it. The observed threat-sensitive latency-to-return response was stronger when the food patch was further away from the nearest refuge. Overall, our results are consistent with the predictions of the threat-sensitive predator avoidance hypothesis in that wild passerine birds （primarily black-capped chickadees Poecile atricapillus） exhibited more intense antipre- dator behavioural responses with increasing level of apparent threat. The birds were thus sensitive to their local perceived threat of predation and traded-off safety from predation （by refuging） and foraging gains in open habitat in a graded, threat-sensitive manner [Current Zoology 60 （6）： 719-728, 2014].
文摘This paper aims to investigate the effects of recasts and working memory on the acquisition of Korean morphological causatives by advanced Chinese learners of Korean. Participants were randomly assigned to two groups： A experimental group and a control group. The experimental group received intensive recasts during task-based interaction, whereas the control group did not. The effects of recasts were measured by two types of tests： An elicited imitation test, as a measure of implicit knowledge, and an untimed grammaticality judgment test, as a measure of explicit knowledge. The findings are as follows. First, from the pretest to the delayed posttest, recasts facilitated the learners＇ acquisition of causative construction by developing both their implicit and explicit knowledge. Second, the results of delayed posttests showed that recasts were more effective in aiding the development of implicit knowledge than explicit knowledge. Third, working memory was found to be a significant covariate in the facilitative effects of recasts on explicit knowledge; that is, it mediated the development of explicit knowledge via recasts as a significant individual difference factor. The effects of recasts were maintained even when the impact of working memory was controlled. （189 words）
