Advancements in language models (LMs) have sparked interest in exploring their potential as knowledge bases (KBs) due to their high capability for storing huge amounts of factual knowledge and semantic understanding. ...
详细信息
In this paper, a multi-media cleaning system based on the optimisation of the tip electrode structure is developed, which significantly improves the electrochemical reaction efficiency and realises the efficient purif...
详细信息
There is a growing interest in expanding the input capacity of language models (LMs) across various domains. However, simply increasing the context window does not guarantee robust performance across diverse long-inpu...
详细信息
In the current era of rapid globalization, companies must focus on their financial operations to capitalize on fast-paced economic development and ensure sustainability. In the financial market, fundamental analysis s...
详细信息
Story video-text alignment, a core task in computational story understanding, aims to align video clips with corresponding sentences in their descriptions. However, progress on the task has been held back by the scarc...
详细信息
The Advanced Metering Infrastructure(AMI),as a crucial subsystem in the smart grid,is responsible for measuring user electricity consumption and plays a vital role in communication between providers and ***,with the a...
详细信息
The Advanced Metering Infrastructure(AMI),as a crucial subsystem in the smart grid,is responsible for measuring user electricity consumption and plays a vital role in communication between providers and ***,with the advancement of information and communication technology,new security and privacy challenges have emerged for *** address these challenges and enhance the security and privacy of user data in the smart grid,a Hierarchical Privacy Protection Model in Advanced Metering Infrastructure based on Cloud and Fog Assistance(HPPM-AMICFA)is proposed in this *** proposed model integrates cloud and fog computing with hierarchical threshold encryption,offering a flexible and efficient privacy protection solution that significantly enhances data security in the smart *** methodology involves setting user protection levels by processing missing data and utilizing fuzzy comprehensive analysis to evaluate user importance,thereby assigning appropriate protection ***,a hierarchical threshold encryption algorithm is developed to provide differentiated protection strategies for fog nodes based on user IDs,ensuring secure aggregation and encryption of user *** results demonstrate that HPPM-AMICFA effectively resists various attack strategies while minimizing time costs,thereby safeguarding user data in the smart grid.
Modern recommendation systems are widely used in modern data *** random and sparse embedding lookup operations are the main performance bottleneck for processing recommendation systems on traditional platforms as they...
详细信息
Modern recommendation systems are widely used in modern data *** random and sparse embedding lookup operations are the main performance bottleneck for processing recommendation systems on traditional platforms as they induce abundant data movements between computing units and ***-based processing-in-memory(PIM)can resolve this problem by processing embedding vectors where they are ***,the embedding table can easily exceed the capacity limit of a monolithic ReRAM-based PIM chip,which induces off-chip accesses that may offset the PIM ***,we deploy the decomposed model on-chip and leverage the high computing efficiency of ReRAM to compensate for the decompression performance *** this paper,we propose ARCHER,a ReRAM-based PIM architecture that implements fully yon-chip recommendations under resource ***,we make a full analysis of the computation pattern and access pattern on the decomposed *** on the computation pattern,we unify the operations of each layer of the decomposed model in multiply-and-accumulate *** on the access observation,we propose a hierarchical mapping schema and a specialized hardware design to maximize resource *** the unified computation and mapping strategy,we can coordinatethe inter-processing elements *** evaluation shows that ARCHER outperforms the state-of-the-art GPU-based DLRM system,the state-of-the-art near-memory processing recommendation system RecNMP,and the ReRAM-based recommendation accelerator REREC by 15.79×,2.21×,and 1.21× in terms of performance and 56.06×,6.45×,and 1.71× in terms of energy savings,respectively.
This research explores multiple varieties of effective and child-friendly learning approaches for mathematics and memorization among young children in today's increasingly digital age considering their specific de...
详细信息
Dynamic multi-mode resource-constrained project scheduling problem (DMRCPSP) is crucial for effectively managing complex projects where activities have multiple options of resource demand and durations are uncertain. ...
详细信息
Adversarial attack for time-series classification model is widely explored and many attack methods are *** there is not a method of attack based on the data *** this paper,we innovatively proposed a black-box sparse a...
详细信息
Adversarial attack for time-series classification model is widely explored and many attack methods are *** there is not a method of attack based on the data *** this paper,we innovatively proposed a black-box sparse attack method based on data *** method directly attack the sensitive points in the time-series data accord-ing to statistical features extract from the *** frst,we have validated the transferability of sensitive points among DNNs with different ***,we use the statistical features extract from the dataset and the sensi-tive rate of each point as the training set to train the predictive ***,predicting the sensitive rate of test set by predictive ***,perturbing according to the sensitive *** attack is limited by constraining the LO norm to achieve one-point *** conduct experiments on several datasets to validate the effectiveness of this method.
暂无评论