Fusion: Practice and Applications

A network Intrusion detection system is a system that can find out different types of attacks. ANIDS is used to find out the noble type of attack by using machine learning and deep learning techniques. These techniques are very useful to find out those attacks whose patterns are not stored in the database. Therefore, these types of systems need more research to improve their accuracy and reduce the false alarm rate. In this paper, we are going to propose an ensemble framework for NIDS using different ML and DL techniques. In this paper, we have used the XGBOOST algorithm for feature extraction and for classification, CNN and RNN deep learning techniques are used. This ensemble model is used for the binary and multiclassification of attacks. Our model was checked on the dataset CICIDS-2018 which gives a better accuracy and low false alarm rate.

This research delves into the developments, in cloud computing and their significance for e government. It introduces an approach to e government advancement known as "Electronic Governmental Cloud (e-GCloud) " aimed at addressing identified issues and meeting the requirements of cloud computing. The study will conduct a review of existing literature and online sources analyzing studies and articles on the evolving landscape of cloud computing to elucidate its role in e government applications. It aims to outline the deployment strategy for cloud computing in e government settings and propose a novel governmental framework called "e- GCloud” designed as an exclusive private cloud community for national governments use. Additionally, this research, evaluates factors influencing the integration of cloud computing into e government systems by drawing insights from senior government officials and IT personnel within governmental entities. The results suggest that e-GCloud outperforms in applications due, to its enhanced flexibility, resource availability and prompt responsiveness.

This research investigates how Egyptian investor sentiment affects cryptocurrency returns, focusing specifically on Bitcoin. We utilized an enhanced investor sentiment index in Egypt, constructed through factor analysis of various literature-based variables. Our study's findings revealed a notable positive correlation between the investor sentiment index, lagged by one order, and Bitcoin returns, as per the estimation and analysis using VAR models. Analysis indicates that a one standard deviation change in the investor sentiment index leads to an alteration in the influence of each standard deviation of the original positive variable, resulting in a switch from positive to negative and vice versa in the medium and long term. Regarding variance decomposition, the short-term variance error of 100% is primarily explained by Bitcoin returns themselves. However, in the medium to long term, besides Bitcoin returns, the investor sentiment index emerges as the most influential variable affecting Bitcoin returns. Causality tests reveal a unidirectional short-term impact from the investor sentiment index to Bitcoin returns via Granger causality tests. Additionally, using the Toda-Yamamoto causality test, long-term bidirectional effects between Bitcoin returns and the investor sentiment index were observed.

The integration of Artificial Intelligence (AI) within the Medical Internet of Things (MIoT) is advancing swiftly, leading to significant developments in the detection of illnesses like epilepsy by analyzing Interictal Epileptiform Discharges (IED) in electroencephalograms (EEG).The availability of EEG data has facilitated the creation of innovative applications, including seizure detection. While neurologists have traditionally relied on EEG data analysis to identify epileptic seizures, the manual evaluation of EEG brain waves is a laborious and complex process that places significant stress on specialists. This paper presents a simple Convolutional Neural Network (CNN) method for the automated detection of IEDs based on EEG waveforms. This approach helps reduce the burden on epilepsy patients by forecasting seizures and enabling timely interventions. It also eases the workload for neurologists and less experienced specialists, thereby accelerating the diagnosis process. The proposed method was implemented by utilizing a series of images that depicted the magnitude of the EEG signal across each sensor. The study divided participants into two groups: (A) healthy individuals and (B) individuals with epilepsy. The results demonstrated an accuracy of up to 96.4% compared to human expert diagnoses, displaying the method's effectiveness and practicality in detecting seizure occurrences in EEG data.

The ability to retrieve a word from the cusp of memory often encounters the well-documented Tip-of-the-Tongue (TOT) barrier. This cognitive phenomenon can impede communication and learning. Addressing this, our study introduces a novel reverse dictionary framework empowered by cutting-edge neural network architectures to facilitate the retrieval of words from definitions or descriptions. This research draws the path of the development and the efficiency of various natural language deep learning models formulated to grasp the semantics inside the text. This work started with gripping a new dataset with rich content from a linguistic perspective. An accurate pre-processing step, including text normalizations and contextual features extraction, was conducted to transform the unstructured text into structured features fitting the model training. Dense vectors representative of text have been extracted using the BERT embedding model. Three models (LSTM, FNN, and GRU) were tested and compared using scrapped and benchmarked data. The proposed model that was consisted from Bert embedding and LSTM learner was evaluated and showed notable performance under cosine similarity and mean square error metrics. The LSTM model proved useful in real-world applications by exhibiting excellent semantic coherence in its embedding and accuracy in its predictions. This research evolved a discussion about the efficient behavior of the pre-trained BERT model in enhancing vocabulary. In addition, this work sheds light on the crucial role of reverse dictionaries in many NLP applications in the future. Subsequent research endeavors will focus on augmenting the multilingual functionalities of our methodology and investigating its suitability for other cognitive linguistic phenomena.