In response to the escalating complexity of cyber threats and the rapid expansion of digital environments, traditional detection models are proving increasingly inadequate. The advent of Large Language Models (LLMs) powered by Natural Language Processing (NLP) represents a transformative advancement in cyber security. This review explores the burgeoning landscape of LLM applications in cyber security, highlighting their significant potential across various threat detection domains. Recent advancements have demonstrated LLMs' efficacy in enhancing tasks such as cyber threat intelligence, phishing detection, anomaly detection through log analysis, and more. By synthesizing recent literature, this paper provides a comprehensive overview of how LLMs are reshaping cyber security frameworks. It also discusses current challenges and future directions, aiming to guide researchers and practitioners in leveraging LLMs effectively to fortify digital defences and mitigate evolving cyber threats

Dynamic malware analysis plays a pivotal role in modern cybersecurity, offering insights into malware behavior through dynamic execution and network traffic analysis. In this study, we present a comprehensive approach to dynamic malware analysis using a sandbox environment and network traffic logs. Our methodology involves the extraction of relevant features from network traffic captured in pcap files. We conducted experiments using a virtualized Oracle VirtualBox environment, where benign and malicious software samples were executed within a Windows virtual machine controlled by Python scripts. For network emulation, we utilized tools from the REMnux distribution, including InetSim and FakeDNS, to simulate realistic network interactions during malware execution. The collected pcap data underwent preprocessing and feature extraction to capture essential behavioral patterns and network indicators. Machine learning and artificial intelligence models were developed to classify malware based on these extracted features. Our findings underscore the efficacy of dynamic analysis coupled with machine learning in detecting and classifying malware variants based on their network behavior. This research contributes to advancing techniques for real-time threat detection and response in cybersecurity, emphasizing the importance of dynamic malware analysis in mitigating evolving cyber threats.

Heart sounds have been extensively studied for heart disease diagnosis for several decades. Traditional machine learning algorithms applied in the literature have typically partitioned heart sounds into small windows and employed feature extraction methods to classify samples. However, as there is no optimal window length that can effectively represent the entire signal, windows may not provide a sufficient representation of the underlying data. To address this issue, this study proposes a novel approach that integrates window-based features with features extracted from the entire signal, thereby improving the overall accuracy of traditional machine learning algorithms. Specifically, feature extraction is carried out using two different time scales. Short-term features are computed from five-second fragments of heart sound instances, whereas long-term features are extracted from the entire signal. The long-term features are combined with the short-term features to create a feature pool known as long short-term features, which is then employed for classification. To evaluate the performance of the proposed method, various traditional machine learning algorithms with various models are applied to the PhysioNet/CinC Challenge 2016 dataset, which is a collection of diverse heart sound data. The experimental results demonstrate that the proposed feature extraction approach increases the accuracy of heart disease diagnosis by nearly 10%.

Alzheimer’s disease is a neurodegenerative condition primarily attributed to environmental factors, abnormal protein deposits, immune system dysregulation, and the consequential death of nerve cells in the brain. On the other hand, Parkinson’s disease manifests as a neurological disorder featuring primary motor, secondary motor, and non-motor symptoms, accompanied by the rapid demise of cells in the brain’s dopamine-producing region. Utilizing brain images for accurate diagnosis and treatment is integral to addressing both conditions. This study harnessed the power of artificial intelligence for classification processes, employing state-of-the-art transformer models such as Swin transformer, vision transformer (ViT), and bidirectional encoder representation from image transformers (BEiT). The investigation utilized an open-source dataset comprising 450 images, evenly distributed among healthy, Alzheimer’s, and Parkinson’s classes. The dataset was meticulously divided, with 80% allocated to the training set (390 images) and 20% to the validation set (90 images). Impressively, the classification accuracy surpassed 80%, showcasing the efficacy of transformer-based models in disease detection. Looking ahead, this study recommends delving into hybrid and ensemble models and leveraging the strengths of multiple transformer-based deep learning architectures. Beyond contributing crucial insights at the intersection of artificial intelligence and neurology, this research emphasizes the transformative potential of advanced models for enhancing diagnostic precision and treatment strategies in Alzheimer’s and Parkinson’s diseases. It signifies a significant step towards integrating cutting-edge technology into mainstream medical practices for improved patient outcomes.

Time series forecasting covers a wide range of topics, such as predicting stock prices, estimating solar wind, estimating the number of scientific papers to be published, etc. Among the machine learning models, in particular, deep learning algorithms are the most used and successful ones. This is why we only focus on deep learning models. Even though it is a hot topic, there are only a few comprehensive studies, and in many studies, there is not much detail about the tested models, which makes it impossible to constitute a comparison chart. Thus, one of the main motivations for this work is to present comprehensive research by providing details about the tested models. In this study, a corpus of the asked questions and their metadata were extracted from the software development and troubleshooting website. Then, univariate time series data were created from the frequency of the questions that included the word “python” as the tag information. In the experiments, deep learning models were trained on the extracted time series, and their prediction performances are presented. Among the tested models, the model using convolutional neural network (CNN) layers in the form of wavenet architecture achieved the best result.

The escalating sophistication of malware poses a formidable security challenge, as it evades traditional protective measures. Static analysis, an initial step in malware investigation, involves code scrutiny without actual execution. One static analysis approach employs the conversion of executable files into image representations, harnessing the potency of deep learning models. Convolutional neural networks (CNNs), particularly adept at image classification, have potential for malware detection. However, their inclination towards structured data requires a preprocessing phase to convert software into image-like formats. This paper outlines a methodology for malware detection that involves applying deep learning models to image-converted executable files. Experimental evaluations have been performed by using CNN models, autoencoder-based models, and pre-trained counterparts, all of which have exhibited commendable performance. Consequently, employing deep learning for image-converted executable analysis emerges as a fitting strategy for the static analysis of software. This research is significant because it utilized the largest dataset to date and encompassed a wide range of deep learning models, many of which have not previously been tested together.