The privacy-preserving nature of Federated Learning (FL) exposes such a distributed learning paradigm to the planting of backdoors with locally corrupted data. We discover that FL backdoors, under a new on-off multi-shot attack form, are essentially stealthy against existing defenses that are built on model statistics and spectral analysis. First-hand observations of such attacks show that the backdoored models are indistinguishable from normal ones w.r.t. both low-level and high-level representations. We thus emphasize that a critical redemption, if not the only, for the tricky stealthiness is reactive tracing and posterior mitigation. A three-step remedy framework is then proposed by exploring the temporal and inferential correlations of models on a trapped sample from an attack. In particular, we use shift ensemble detection and co-occurrence analysis for adversary identification, and repair the model via malicious ingredients removal under theoretical error guarantee. Extensive experiments on various backdoor settings demonstrate that our framework can achieve accuracy on attack round identification of ∼80% and on attackers of ∼50%, which are ∼28.76% better than existing proactive defenses. Meanwhile, it can successfully eliminate the influence of backdoors with only a 5%∼6% performance drop.
By using efficient and timely medical diagnostic decision making, clinicians can positively impact the quality and cost of medical care. However, the high similarity of clinical manifestations between diseases and the limitation of clinicians’ knowledge both bring much difficulty to decision making in diagnosis. Therefore, building a decision support system that can assist medical staff in diagnosing and treating diseases has lately received growing attentions in the medical domain. In this paper, we employ a multi-label classification framework to classify the Chinese electronic medical records to establish corresponding relation between the medical records and disease categories, and compare this method with the traditional medical expert system to verify the performance. To select the best subset of patient features, we propose a feature selection method based on the composition and distribution of symptoms in electronic medical records and compare it with the traditional feature selection methods such as chi-square test. We evaluate the feature selection methods and diagnostic models from two aspects, false negative rate (FNR) and accuracy. Extensive experiments have conducted on a real-world Chinese electronic medical record database. The evaluation results demonstrate that our proposed feature selection method can improve the accuracy and reduce the FNR compare to the traditional feature selection methods, and the multi-label classification framework have better accuracy and lower FNR than the traditional expert system.
1.Dorysthenes paradoxus,including the adults,pupae and larvae,were collected in Swei Chung (1964 Wang and Feng) and Zhuanghe (1974 Wang and Li).We dissected them all and found the juveniles of Macracanthorhynchus hirudinaceus in each of their bodies.It is the first discovery of D.paradoxes as the intermediate host of the worm.In Chuang Ho the incidence of the infection of M.hirudinaceus among 165 adult cerambyeids is 38.8% (9% among 25 females and 55% among 140 males),while in Chen Shan commune the incidence is as high as 62.5% (20% among 35 pupae and 15% among 100 larvae).Generally there is only one single juvenile in the body of a host of D.paradoxus.But sometimes there are as many as 27-40 and in average 4.1 juvenils contained in a host of D.paradoxus.2.The juvenile worm has been measured and described in detail.The juvenile is also called cyst acanthella,with a size of 2.4-3.9 mm×1.6-2.0 mm and milky white in colour.The anterior is wider and flatter with a size 0.7-0.9 mm.An imagination is in middle.The proboscis is retracted into the sheeth somewhat like a sesame seed.The booklet is evidently visible.The posterior part is more narrow with a size of 0.2-0.6 mm and in the 1/5 of its posterior part 7-8 transverse figures are seen.Body wall is 0.2-0.3 mm in thickness.Two lemnisci are of strip shape and are situated on both sides of the proboscis.In the middle of the body there is a longitudinal ligament from posterior to the end of proboscis.The reproductive organs of both the male and the female are not clearly visible.3.The morphology and life habit of D.paradoxus have also been observed.4.Human infection of M.hirudinaceus is chiefly due to the habit of eating raw or half-raw D.paradoxus.It has been verified that D.paradoxus is the intermediate host of M.hindinaceus in Liaoning Province.
Telemedicine is expected to play a significant role in previsioning better service and meeting various needs of patients, healthcare providers, and policy makers by exploiting advanced information and communication technologies. Because health data are highly sensitive, the security and privacy of telemedicine should be protected carefully to protect telemedicine from being potentially interrupted and attacked. In telemedicine networks, the sink node is the most vulnerable to attacks because it is where sensitive data will be collected. Thus, it is of great importance to protect the privacy of the sink node in telemedicine networks. To this end, this paper proposes an efficient privacy-preserving protocol for sink node location in telemedicine networks. Compared with the existing work, the scheme can improve the safe time of telemedicine networks by injecting request packets and reduce the delivery time by transmitting along the shortest route. In particular, the safe time can be improved by 28.57% to 52.70%, and the delivery time can be reduced by 22.86% to 27.61%.
Risk probability estimating analyzes the distributed network and assesses inherently uncertain events and circumstances dynamically.And the estimating probability of security risk is an absolutely necessary step of developing network security protection system.The problems of feature extraction,cluster analysis,similarity measurement and estimation methods in network situation awareness were addressed.A risk probability assessment formula was proposed,and an estimating model adopting the neural networks was presented.An experiment based on DARPA intrusion detection evaluation data was given to support the suggested approach and demonstrated the feasibility and suitability for use.
With pure normal training videos, video abnormal event detection (VAD) aims to build a normality model, and then detect abnormal events that deviate from this model. Despite of some progress, existing VAD methods typically train the normality model by a low-level learning objective (e.g. pixel-wise reconstruction/prediction), which often overlooks the high-level semantics in videos. To better exploit high-level semantics for VAD, we propose a novel paradigm that performs VAD by learning a Consistency-Aware high-level Feature Extractor (CAFE). Specifically, with a pre-trained deep neural network (DNN) as teacher network, we first feed raw video events into the teacher network and extract the outputs of multiple hidden layers as their high-level features, which contain rich high-level semantics. Guided by high-level features extracted from normal training videos, we train a student network to be the high-level feature extractor of normal events, so as to explicitly consider high-level semantics in training. For inference, a video event can be viewed as normal if the student extractor produces similar high-level features to the teacher network. Second, based on the fact that consecutive video frames usually enjoy minor differences, we propose a consistency-aware scheme that requires high-level features extracted from neighboring frames to be consistent. Our consistency-aware scheme not only encourages the student extractor to ignore low-level differences and capture more high-level semantics, but also enables better anomaly scoring. Last, we also design a generic framework that can bridge high-level and low-level learning in VAD to further ameliorate VAD performance. By flexibly embedding one or more low-level learning objectives into CAFE, the framework makes it possible to combine the strengths of both high-level and low-level learning. The proposed method attains state-of-the-art results on commonly-used benchmark datasets.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)
'%3e%3ccircle r='20' fill='%23008'/%3e%3ccircle r='17.5' fill='%23fff'/%3e%3ccircle r='3.5' fill='%23008'/%3e%3cg id='d'%3e%3cg id='c'%3e%3cg id='b'%3e%3cg id='a' fill='%23008'%3e%3ccircle r='.875' transform='rotate(7.5 -8.75 133.5)'/%3e%3cpath d='M0 17.5.6 7 0 2l-.6 5L0 17.5z'/%3e%3c/g%3e%3cuse xlink:href='%23a' transform='rotate(15)'/%3e%3c/g%3e%3cuse xlink:href='%23b' transform='rotate(30)'/%3e%3c/g%3e%3cuse xlink:href='%23c' transform='rotate(60)'/%3e%3c/g%3e%3cuse xlink:href='%23d' transform='rotate(120)'/%3e%3cuse xlink:href='%23d' transform='rotate(-120)'/%3e%3c/g%3e%3c/svg%3e)