Malicious applications are usually comprehended through two major techniques, namely static and dynamic analyses. Through static analysis, a given malicious program is parsed, and some representative artifacts (e.g., control-flow graphs) are produced without any execution; whereas, the given malicious application needs to be executed when conducting dynamic analysis. These two mainstream techniques for analyzing the given software are effective in detecting certain classes of malware. More specifically, through static analysis, the patterns and signature of the malware are exposed, helping in detecting any known malicious payload hidden in or injected into the code. On the other hand, behavioral and run-time execution patterns of software are explored through dynamic analysis. To ease the analysis process, a third analysis approach, known as the visual representation of the artifacts created by both static and dynamic analysis tools, would also be a supplementary asset for malware experts. This paper introduces “ MalView ”, an interactive visualization platform, for malware analysis by which pattern matching techniques on both signature-based and behavioral analysis artifacts can be utilized to 1) classify malware, 2) identify the intention and location of the malicious payload in the artifacts, 3) analyze unknown malware (i.e., zero-day malware) by recognizing any unusual signature or behavior, and 4) explore the time dependencies and thus the system components affected or tampered by the underlying malware. The results of several case studies conducted in this work show that MalView offers more features and information compared to some other visualization tools, facilitating the malware analysis process.

The existing literature on forecasting time series data is primarily based on univariate analysis and techniques such as Univariate Autoregressive (UAR), Univariate Moving Average (UMA), Simple Exponential Smoothing (SES), deep learning models, and, most notably, univariate Long Short-Term Memory (LSTM) built based on univariate variable where the next lag of time series is leveraged for forecasting the next cycle of data. This paper takes this line of research to the next level by focusing on forecasting time series data based on “multivariate” modeling and analysis. To have a better insight of the performance of various deep learning-based models when multivariate analysis is performed, the paper builds and reports the forecasting accuracy for techniques such as the Transformer-based Multi-head Attention network, Long Short-Term Memory (LSTM), Bidirectional Long Short-Term Memory (BI-LSTM), Temporal Convolution Network (TCN), and conventional Vector Autoregressive (VAR) models. The findings revealed that the TCN model achieved the average lowest RMSE values of 0.0589 for stock data and 0.1554 for cryptocurrency data. Notably, the Multi-Head Attention model achieved average R ² values of 0.92 for stock data and −1.98 for cryptocurrency data with respect to five variables (i.e., open, high, low, close and volume). According to the empirical studies conducted and reported in this paper, the transformer-based Multi-head Attention network outperformed other models such as LSTM, BI-LSTM, and more importantly conventional Vector Auto-Regression Models (VAR) in stocks and cryptocurrencies time series data where several variables were leveraged in building these multivariate-based models.

Cyber attacks continue to pose significant threats to individuals and organizations, stealing sensitive data such as personally identifiable information, financial information, and login credentials. Hence, detecting malicious websites before they cause any harm is critical to preventing fraud and monetary loss. To address the increasing number of phishing attacks, protective mechanisms must be highly responsive, adaptive, and scalable. Fortunately, advances in the field of machine learning, coupled with access to vast amounts of data, have led to the adoption of various deep learning models for timely detection of these cyber crimes. This study focuses on the detection of phishing websites using deep learning models such as Multi-Head Attention, Temporal Convolutional Network (TCN), BI-LSTM, and LSTM where URLs of the phishing websites are treated as a sequence. The results demonstrate that Multi-Head Attention and BI-LSTM model outperform some other deep learning-based algorithms such as TCN and LSTM in producing better precision, recall, and F1-scores.

It is important to comprehend the attacker’s behavior and capacity in order to build a stronger fortress and thus be able to protect valuable assets more effectively. Prior to launching technical and physical attacks, an attacker may enter the reconnaissance stage and gather sensitive information. To collect such valuable data, one of the most effective approaches is through conducting social engineering attacks, borrowing techniques from deception theory. As a result, it is of utmost importance to understand when an attacker behaves truthfully and when the attacker opts to be deceitful. This paper models attacker’s states using Markov Decision Process (MDP) and studies the attacker’s decision for launching deception attacks in terms of cooperation and deception costs. The study is performed through Markov Decision Process (MDP) modeling, where the states of attackers are modeled along with the permissible actions that can be taken. We found that the optimal policy regarding being deceitful or truthful depends on the cost associated with deception and how much the attacker can afford to take the risk of launching deception attacks. More specifically, we observed that when the cost of cooperation is low (e.g., 10%), by taking MDP optimal policy, the attacker cooperates with the victim as much as possible in order to gain their trust; whereas, when the cost of cooperation is high (e.g., 50%), the attacker takes deceptive action earlier in order to minimize the cost of interactions while maximizing the impact of the attack. We report four case studies and simulations through which we demonstrate the trade-off between cooperative and deceptive actions in accordance with their costs to attackers.

Sonification is the utilization of sounds to convey information about data or events. There are two types of emotions associated with sounds: (1) “perceived” emotions, in which listeners recognize the emotions expressed by the sound, and (2) “induced” emotions, in which listeners feel emotions induced by the sound. Although listeners may widely agree on the perceived emotion for a given sound, they often do not agree about the induced emotion of a given sound, so it is difficult to model induced emotions. This paper describes the development of several machine and deep learning models that predict the perceived and induced emotions associated with certain sounds, and it analyzes and compares the accuracy of those predictions. The results revealed that models built for predicting perceived emotions are more accurate than ones built for predicting induced emotions. However, the gap in predictive power between such models can be narrowed substantially through the optimization of the machine and deep learning models. This research has several applications in automated configurations of hardware devices and their integration with software components in the context of the Internet of Things, for which security is of utmost importance.

In machine learning class imbalance is a pressing issue, where the model is biased towards the majority classes and underperforms in the minority classes. In textual data, the natural language processing (NLP) model bias significantly reduces overall accuracy, along with poor performance in minority classes. This paper investigates and compares the performance of transformer-based models, such as Multi-head Attention with the data levels and algorithmic levels approaches and BERT (Bidirectional Encoder Representations from Transformers) with LLM-based data augmentation. The research utilized the approaches, such as Random Over Sampler, Synthetic Minority Over-sampling Technique (SMOTE), SMOTEENN, data augmentation at word level, class weights, L2 regularization and leveraging GPT-3.5-Turbo’s for data augmentation to create additional data samples in imbalance dataset. The results from the experiment demonstrate that the LLM-based data augmentation with Multi-head Attention and BERT in the Myers-Briggs Type Indicator (MBTI) dataset (a highly skewed dataset) achieves the highest precision, recall and F1 score of 0.76 across terms. It indicates that the LLM-based data augmentation has significant improvements in dealing with class imbalance and improves the model’s accuracy in minority class types in the MBTI dataset.