This study presents a smart doorbell system designed to improve home security through advanced deep learning techniques. At its core, the system utilizes the YOLOv5 (You Only Look Once) algorithm for real-time face de...
详细信息
Prediction of high performance concrete (HPC) compressive strength is very important for determination of mix design optimization and structural reliability. An integrated deep learning framework, together with a Baye...
详细信息
Continual Learning (CL) faces challenges such as catastrophic forgetting and data scarcity in low-resource tasks, which limit the model's ability to adapt to gradually changing tasks. We propose a general CL metho...
详细信息
ISBN:
(纸本)9798400713316
Continual Learning (CL) faces challenges such as catastrophic forgetting and data scarcity in low-resource tasks, which limit the model's ability to adapt to gradually changing tasks. We propose a general CL method for low-resource tasks to address these challenges called GAIN. This method addresses cross-domain continual learning through two approaches: (i) the Gradual Adaptation Module, which incrementally stacks lightweight adapters to effectively retain knowledge from previous tasks and adapt to new ones, thereby mitigating catastrophic forgetting, and (ii) a bottleneck layer size tuning strategy to improve learning efficiency in low-resource scenarios. Finally, we have extensive experiments on four datasets to validate the effectiveness and robustness of GAIN, showing that it alleviates catastrophic forgetting and enhances learning efficiency across various task types.
Congestion is a prevalent challenge to every networks today due to the higher network access and inefficient utilization of available bandwidth. Bottleneck Bandwidth and Round-trip propagation time (BBR) is a popular ...
详细信息
Word Sense Disambiguation(WSD) is a Natural Language Processing(NLP) technique that tries to disambiguate ambiguous words by finding the right sense of a word in a particular context. It is a classification task that ...
详细信息
ISBN:
(数字)9798331512248
ISBN:
(纸本)9798331512255
Word Sense Disambiguation(WSD) is a Natural Language Processing(NLP) technique that tries to disambiguate ambiguous words by finding the right sense of a word in a particular context. It is a classification task that assists in mapping a word to one or more potential senses based on its sentence. Understanding the intended sense in a particular context is crucial for proper communication and accurate language processing. But WSD is especially crucial in Bengali for a number of reasons: Similar to other Indic languages, Bengali possesses a rich and complex system of word formation. A root word can be modified according to aspects like tense, gender, number, and case. An attempt is made in this working paper where authors will be working on WSD but on the Bengali language, where words have different meanings. The Bengali language has plenty of words with multiple meanings. Even though WSD has been propagated the world over, the language Bengali has a lot of challenges since there are very few resources available and research due to its complex syllable structure.
In this letter, we present the latest findings in coexistence of two rather different classes of optical signals in one fiber - classical data signals and quantum signals. The most significant and notable difference b...
详细信息
Recently, social media platforms have become very popular as they offer unbelievable opportunities to their users. Twitter is one of the social media platforms on which a huge number of people exchange their messages ...
详细信息
Recently, social media platforms have become very popular as they offer unbelievable opportunities to their users. Twitter is one of the social media platforms on which a huge number of people exchange their messages by posting tweets. However, this platform is usually used by automated accounts called bots. Such bots are used to spread fake news, fake ideas, and products. Hence, it is essential to detect the presence of spam bots on Twitter. In order to detect spam bots on Twitter, an effective feature selection technique using a novel hybrid deep learning model is introduced in this paper. This paper proposes a novel spam bot detection system for the Twitter social network that combines profile and tweet-based features. Initially, the Twitter data are pre-processed to improve the accuracy of classification. The pre-processing stage involves various steps such as stopping word removal, tokenization, stemming, n-gram identification, user mention, and vocabulary density and richness. After pre-processing, the tweets are given to the next stage for feature extraction. In this stage, the user profile-based features such as name, screen name, location, and time, as well as the tweet-based features such as hashtags, retweeting of tweets, etc., are extracted from the tweets. The extracted features are then subjected to feature selection, where a meta-heuristic-based optimization algorithm called the Binary Golden Search Optimization algorithm (BGSO) is used. This method helps to reduce the feature dimensionality and overfitting issues. In order to improve the optimization algorithm’s searching ability, an X-shaped transfer function is used. Finally, the selected features are provided to the novel Hybrid Hopfield Dilated Depthwise Separable Convolutional Neural Network (HHD 2 SCNN) based classification model, where the output layer classifies the given tweets as spam bots or legitimate. The proposed method is experimentally verified, and the performance metrics are evaluat
The presence of numerous bidirectional communication devices connecting customers to the grid makes smart grid networks particularly vulnerable to network attacks. As an instance of a network attack that could impact ...
详细信息
ISBN:
(数字)9798331509675
ISBN:
(纸本)9798331509682
The presence of numerous bidirectional communication devices connecting customers to the grid makes smart grid networks particularly vulnerable to network attacks. As an instance of a network attack that could impact the smart grid, a distributed denial of service (DDoS) attack occurs when a compromised grid communiqué devices or nodes send a flood of false data or requests to the smart grid network. This can cause problems with smart meters, data servers, and the state estimator, which in turn affects the services that end users receive. The problem of protecting the network from distributed denial of service (DDoS) attacks is best addressed by methods that rely on machine learning. One major drawback of using machine learning-based approaches is that models need to be retrained every time new types of attacks appear. Despite the fact that DDoS attack detection has long relied on Intrusion Detection Systems (IDS), previous research has been limited due to a number of factors. Actually, it is highly discouraged to interfere with the regular functioning of the smart grid. to suggest using reconstructive deep learning methods to identify DDoS attacks efficiently and with minimal downtime in order to tackle this difficulty. In order to identify typical attacks on a communication network, this research introduces a shallow neural network with just 110 artificial neurones triggered by ReLU. to suggest a better attack-sharing loss function to handle unbalanced learning, which will allow for the training of such tiny neural networks. A Lungs Performance-based Optimisation (LPO) was devised using the suggested model to hone the parameters. In smart grid settings, our suggested approach improves the scalability and dependability of IDS and provides useful insights for building strong IDS based on machine learning. This enables organisations to strengthen their defences against harmful cyber-attacks and proactively reduce security risks.
The proper categorization of date fruits based on their genetic variations is a crucial component of effective crop management and quality control in the field of agriculture. Nonetheless, the present methodologies ut...
详细信息
Speculative decoding has been shown as an effective way to accelerate Large Language Model (LLM) inference by using a Small Speculative Model (SSM) to generate candidate tokens in a so-called speculation phase, which ...
详细信息
暂无评论