Recent Submissions

  • ViTDroid: Vision Transformers for Efficient, Explainable Attention to Malicious Behavior in Android Binaries

    Syed, Toqeer Ali; Nauman, Mohammad; khan, Sohail; Jan, Salman; Zuhairi, Megat; External Collaboration; Artificial Intelligence & Cyber Security Lab; 0; 0; Computer Science; et al. (MDPI, 2024-10-17)
    Smartphones are intricately connected to the modern society. The two widely used mobile phone operating systems, iOS and Android, profoundly affect the lives of millions of people. Android presently holds a market share of close to 71% among these two. As a result, if personal information is not securely protected, it is at tremendous risk. On the other hand, mobile malware has seen a year-on-year increase of more than 42% globally in 2022 mid-year. Any group of human professionals would have a very tough time detecting and removing all of this malware. For this reason, deep learning in particular has been used recently to overcome this problem. Deep learning models, however, were primarily created for picture analysis. Despite the fact that these models have shown promising findings in the field of vision, it has been challenging to fully comprehend what the characteristics recovered by deep learning models are in the area of malware. Furthermore, the actual potential of deep learning for malware analysis has not yet been fully realized due to the translation invariance trait of well-known models based on CNN. In this paper, we present ViTDroid, a novel model based on vision transformers for the deep learning-based analysis of opcode sequences of Android malware samples from large real-world datasets. We have been able to achieve a false positive rate of 0.0019 as compared to the previous best of 0.0021. However, this incremental improvement is not the major contribution of our work. Our model aims to make explainable predictions, i.e., it not only performs the classification of malware with high accuracy, but it also provides insights into the reasons for this classification. The model is able to pinpoint the malicious behavior-causing instructions in the malware samples. This means that our model can actually aid in the field of malware analysis itself by providing insights to human experts, thus leading to further improvements in this field.
  • Metabolomic profiling reveals altered phenylalanine metabolism in Parkinson’s disease in an Egyptian cohort

    Sheb, Nourhan; El-Jaafary, Shaimaa; A. Saeed, Ayman; ElKafrawy, Passent; El-Sayed, Amr; Shamma, Samir; Elnem, Rasha; Mekky, Jaidaa; Mohamed, Lobna A.; Kittaneh, Omar; et al. (Frontiers, 2024-03-07)
    Introduction: Parkinson’s disease (PD) is the most common motor neurodegenerative disease worldwide. Given the complexity of PD etiology and the different metabolic derangements correlated to the disease, metabolomics profiling of patients is a helpful tool to identify patho-mechanistic pathways for the disease development. Dopamine metabolism has been the target of several previous studies, of which some have reported lower phenylalanine and tyrosine levels in PD patients compared to controls.Methods: In this study, we have collected plasma from 27 PD patients, 18 reference controls, and 8 high-risk controls to perform a metabolomic study using liquid chromatography-electrospray ionization–tandem mass spectrometry (LC-ESI-MS/MS).Results: Our findings revealed higher intensities of trans-cinnamate, a phenylalanine metabolite, in patients compared to reference controls. Thus, we hypothesize that phenylalanine metabolism has been shifted to produce trans-cinnamate via L-phenylalanine ammonia lyase (PAL), instead of producing tyrosine, a dopamine precursor, via phenylalanine hydroxylase (PAH).Discussion: Given that these metabolites are precursors to several other metabolic pathways, the intensities of many metabolites such as dopamine, norepinephrine, and 3-hydroxyanthranilic acid, which connects phenylalanine metabolism to that of tryptophan, have been altered. Consequently, and in respect to Metabolic Control Analysis (MCA) theory, the levels of tryptophan metabolites have also been altered. Some of these metabolites are tryptamine, melatonin, and nicotinamide. Thus, we assume that these alterations could contribute to the dopaminergic, adrenergic, and serotonergic neurodegeneration that happen in the disease.
  • Promoting Sales of Knowledge Products on Knowledge Payment Platforms: A LargeScale Study with a Machine Learning Approach

    Zhang, Jacky; Jiang, Shan; Wang, Xuyan; Duan, Keran; Xiao, Yuting; Lytras, Miltiadis; Xu, Dongming; Zheng, Yunhao; Ordonez De Pablos, Patricia; External Collaboration; et al. (Elsevier, 2024-05-14)
    With the digital transformation of the global economy, a new mode of knowledge service has emerged on open innovation platforms, such as those for the sharing economy. This mode is the paid knowledge-sharing service, where knowledge providers share knowledge only with those who have paid for it. Since an individual customer’s purchases are influenced by others around them, we adopted social influence theory to explain sales of such services on paid knowledge-sharing platforms. A machine learning approach was applied to analyze 27,223 text reviews from the Zhihu Live platform, a well-known and large-scale open knowledge community in China. Hierarchical regression models were built to verify twelve proposed hypotheses about the knowledge providers, knowledge quality, interaction quality, and ratings. The results confirm the positive effect on sales of responsiveness, a dimension of interaction quality, and the negative effect on sales of free provider-driven knowledge contributions. In summary, this study provides a comprehensive framework for antecedent factors of sales of knowledge-sharing services. By introducing knowledge management notions from the field of e-commerce (e.g., price, quality), this study broadens the understanding of the free-to-paid phenomenon on knowledge-sharing platforms.
  • An Ensemble Voting Approach With Innovative Multi-Domain Feature Fusion for Neonatal Sleep Stratification

    Subasi, Abdulhamit; Siddiqa, Hafza Ayeshaa; Nahliis, Abdelwahed; Chen, Chen; Xu, Yan; Wang, Laishuan; Nawaz, Anum; Westerlund, Tomi; Chen, Wei; External Collaboration; et al. (IEEE, 2023-12)
    A limited number of electroencephalography (EEG) channels are useful for neonatal sleep classification, particularly in the Internet of Medical Things (IoMT) field, where compact and lightweight devices are essential to monitoring health effectively. A streamlined and cost-effective IoMT solution can be achieved by utilizing fewer EEG channels, thereby reducing data transmission and device processing requirements. Using only two channels of an EEG device, this study presents a binary and multistage classification of neonatal sleep. The binary classification (sleep vs awake) achieved an accuracy of 87.56%, and a Cohen’s kappa of 74.13%. The quiet sleep ( QS ) detection accuracy was 95.63%, with a Cohen’s kappa of 83.87%. For the three-stage classification, accuracy was 83.72%, and Cohen’s kappa was 69.73%. With only two channels, these are the highest performance parameters. The focus is on the fusion of features extracted through flexible analytical wavelet transform (FAWT) & discrete wavelet transform (DWT), ensemble-based voting models, and fewer channels. To feed crucial features into the ensemble-based voting model, feature importance, feature selection, and validation mechanisms were used. To design the voting classifier, several machine learning models were used, compared, and optimized. With SelectKBest feature selection, the proposed methodology was found to be the most effective. By using only two channels, this study shows the practicality of classifying neonatal sleep stages.
  • An IoT-Based Non-Contact ECG System: Sole of the Feet / Hands Palm

    Subasi, Abdulhamit; External Collaboration; Artificial Intelligence & Cyber Security Lab; 0; 0; Computer Science; 1; Irfan, M. (IEEE, 2023-11)
    In smart healthcare facilities designed especially for the elderly, noncontact electrocardiogram (ECG) measurements could provide essential information about an elderly person’s health by enabling long-term health analytics. In this research work, we propose an Internet of Things (IoT)-based noncontact ECG measurement system. The noncontact measurement is done using flexible electrodes that are made of fabric. These fabric-based flexible electrodes are designed to measure ECG signals from the sole of the feet (SOF) or the palms of the hands (POHs) without touching human skin. To mitigate the impact of nearby electromagnetic radiation on the electrodes, a double layer of isopotential shielding is placed underneath the two active electrodes. The gathered biosignals are stored in the IoT device and transmitted to the cloud. To reduce the amount of stored and transmitted data, we improved our adaptive coding algorithm. The adaptive coding results in an average data reduction of 72%. The data can be fully recovered in the cloud for further analyses using advanced cloud-based tools in ThingSpeak. The study tested the proposed system on 35 participants, including elderly persons, adults, and children. Based on the experiments, the proposed system accurately measures the ECG signal. We validated the results with the ground truth data [polysomnography
  • Automated Facial Expression Recognition using Novel Textural Transformation

    Dogan, S; Subasi, Abdulhamit; External Collaboration; Artificial Intelligence & Cyber Security Lab; 0; 0; Computer Science; 0; Tuncer, T. (Springer, 2023-04)
    Facial expressions demonstrate the important information about our emotions and show the real intentions. In this study, a novel texture transformation method using graph structures is presented for facial expression recognition. Our proposed method consists of five steps. First the face image is segmented and resized. Then the proposed graph-based texture transformation is used as feature extractor. The exemplar feature extraction is performed using the proposed deep graph texture transformation. The extracted features are concatenated to obtain one dimensional feature set. This feature set is subjected to maximum pooling and principle component analysis methods to reduce the number of features. These reduced features are fed to classifiers and we have obtained the highest classification accuracy of 97.09% and 99.25% for JAFFE and TFEID datasets respectively Moreover, we have used CK + dataset to obtain comparison results and our textural transformation based model yielded 100% classification accuracy on the CK + dataset. The proposed method has the potential to be employed for security applications like counter terrorism, day care, residential security, ATM machine and voter verification.
  • OzNet: A New Deep Learning Approach for Automated Classification of COVID-19 Computed Tomography Scans

    Ozgur, Yeniay; Subasi, Abdulhamit; Ozaltin, Oznur; External Collaboration; Artificial Intelligence & Cyber Security Lab; 0; 0; Computer Science; 1; Ozaltin, Oznur (2023-12)
    Coronavirus disease 2019 (COVID-19) is spreading rapidly around the world. Therefore, the classification of computed tomography (CT) scans alleviates the workload of experts, whose workload increased considerably during the pandemic. Convolutional neural network (CNN) architectures are successful for the classification of medical images. In this study, we have developed a new deep CNN architecture called OzNet. Moreover, we have compared it with pretrained architectures namely AlexNet, DenseNet201, GoogleNet, NASNetMobile, ResNet-50, SqueezeNet, and VGG-16. In addition, we have compared the classification success of three preprocessing methods with raw CT scans. We have not only classified the raw CT scans, but also have performed the classification with three different preprocessing methods, which are discrete wavelet transform (DWT), intensity adjustment, and gray to color red, green, blue image conversion on the data sets. Furthermore, it is known that the architecture's performance increases with the use of DWT preprocessing method rather than using the raw data set. The results are extremely promising with the CNN algorithms using the COVID-19 CT scans processed with the DWT. The proposed DWT-OzNet has achieved a high classification performance of more than 98.8% for each calculated metric.
  • Feature extraction and fusion using deep convolutional neural networks for Covid-19 detection using CT and X-RAY images

    Keles, T.; Ozyurt, F; Dogan, S; Tuncer, T.; Subasi, Abdulhamit; External Collaboration; Artificial Intelligence & Cyber Security Lab; 0; 0; Computer Science; et al. (2023)
    COVID-19 represents a novel variant of the coronavirus disease, having rapidly disseminated across the globe. In recent studies pertaining to computer vision, image processing, and classification techniques, numerous methodologies have been introduced employing chest X-ray images and computerized tomography (CT) images. This investigation introduces novel fused deep features founded on CT and X-ray image classification, with the aim of detecting COVID-19 disease. The proposed approach encompasses three key phases: deep feature generation, iterative feature selection, and classification. During the feature generation phase, well-known pre-trained deep convolutional neural networks, namely DenseNet201, MobileNetV2, ResNet18, ResNet50, ResNet101, VGG16, and VGG19, were leveraged. Each network model generated 1000 features, which were subsequently fused, culminating in the acquisition of a final feature vector comprising 7000 dimensions. In order to distill the most pertinent information, the ReliefF and iterative maximum relevance minimum redundancy (RFImRMR) feature selection techniques were employed in the generation of the ultimate feature vector. To evaluate the performance of the proposed method, publicly available datasets of CT images and X-ray images were employed. Notably, the suggested deep learning approach attained accuracy rates of 99.33% and 93.10% for COVID-19 detection using CT and X-ray images, respectively. These achieved results serve as compelling evidence substantiating the efficacy of the proposed fused deep features and RFImRMR-based COVID-19 detection.
  • Tractable Executable Binary Provenance Signalling through Vision Transformers

    Nauman, Mohammad; No Collaboration; Artificial Intelligence & Cyber Security Lab; 0; 0; Computer Science; 0; Nauman, Mohammad (2024-01-15)
    Provenance signaling involves tracing the source information of digital artifacts. It is a valuable intermediate output that greatly facilitates upstream tasks, including but not limited to malware analysis. Existing approaches to provenance signaling either rely on fully manual analysis or machine learning-based models that heavily depend on manually curated input features. This curation process requires the involvement of human experts, which is not only time-consuming but also infeasible on a large scale. In this paper, we present a novel model for provenance signaling that takes raw binaries as input and provides provenance signals with high efficacy. Our model is based on the state-of-the-art vision transformer architecture. We create a novel pipeline of efficiently encoding any binary into 2D sequences, capturing large-scale spatial relations hidden among binary opcodes. This allows our model to extract meaningful information about provenance without requiring the involvement of a human expert. Therefore, our work produces high-accuracy results and provides insights into the learning process, thus making the results more explainable.
  • Clustering potential metaverse users with the use of a value-based framework: Exploiting perceptions and attitudes on the use and adoption of metaverse for bold propositions

    Rsha Mirza; Lytras, Miltiadis; Ohoud Alzamzami; Lama Al Khuzayem; Hajar Alharbi; Sultanah Alshammari; Alaa Bafail; Arwa Basbrain; Eaman Alharbi; Nada Bajnaid; et al. (Pergamon, 2024-03-01)
    Metaverse is a new emerging platform that enables users to interact with each other and engage in many activities inside a collective virtual shared space. The metaverse is rapidly evolving, and it is essential to understand the perception and attitude of its potential users. Therefore, this research aims to study the users of the metaverse and obtain their opinions on some of the key variables of the metaverse. These variables include the core metaverse concept, readiness, ease of use, intention to belong to metaverse, intention to adopt, and value realization. We designed an online questionnaire that aims to measure these variables. Then, a deep analysis of the collected data was conducted using various exploratory descriptive statistics on all participants. Subsequently, we performed further descriptive and correlation analysis on the participants from Saudi Arabia, which resulted in clustering participants into three groups. The result of this work can help in understanding the current and potential users of the metaverse, especially the users from Saudi Arabia. This study will consequently help in developing the metaverse space, enhancing its features, and providing users with the best experiences. Our research contributes to the theory of the metaverse by justifying a bold value-based framework for metaverse adoption. This study also introduces three clusters, Skeptical, Unaware, and Optimists, of potential users of metaverse platforms, providing a clear description of each group.
  • An improved clustering method using particle swarm optimization algorithm and mitochondrial fusion model (PSO-MFM)

    Elkafrawy, Passent; Nasef, Mohamed; Hashim, Amal; External Collaboration; Virtual Reality Lab; 0; 0; Computer Science; 1; Nasef, Mohamed (IOS Press BV, 2024-02-16)
    Computational models are foundational concepts in computer science; many of these models such as P systems are based on natural biological processes. P systems represent a wide framework for a variety of concepts of data mining, as models of data clustering approaches. Data clustering is a technique for analyzing data based on its structure that is widely utilized for many applications. In this paper, the proposed model (PSO-MFM) has combined the Particle Swarm Optimization algorithm (PSO) with Mitochondrial Fusion Model to overcome some constraints of clustering techniques. The solving of clustering problem based on particle swarm is investigated in the proposed model when mutual dynamic rules are used. It can find the best cluster centers for a data set and improve clustering performance by utilizing the distributed parallel computing concept of mutual dynamic rules of mitochondrial fusion model. The comparative results demonstrate that the proposed strategy outperforms competition models when it comes to clustering accuracy, stability and the most efficient in time complexity.
  • Advancing precision medicine in medical education: Integrated, precise and data-driven smart solutions

    Paraskevi Papadopoulou; Lytras, Miltiadis; External Collaboration; NA; NA; NA; Computer Science; NA; Paraskevi Papadopoulou (2023-12-02)
    Advances in “Precision Medicine” initiative, also known as “Personalized Medicine” is an emerging approach for disease treatment and prevention that has already led to innovative discoveries and has created “smart” applications and solutions tailored to a person's or a group of individuals' genetic profile, lifestyle, and environment interaction. Already many physicians as part of patient care routinely prescribe various molecular/genetic and other tests enabling them to select personalized treatments that improve the chances of survival and reduce exposure to adverse effects. This initiative should provide to medical and healthcare professionals with adequate resources and readily available solutions so that the target to specific treatments and care of the illnesses is achieved while at the same time protecting the privacy and safety of the individual is secured as well as the Electronic Health Records (EHRs) and whatever additional data is necessary within the context of Precision Medicine. This study, through a literature review mostly, and some case study analysis, examine whether personalized medicine is delivered to the patient in an “accurate” and “precise” way, as expected. This requires that Health and Human Services and other stakeholders and agencies collaborate to solicit the right input from patients while at the same time can identify and address any educational, practical, legal, and technical issues and providing smart solutions. The lack of proper medical education and advanced infrastructure are still major barriers to the adoption of Precision Medicine, therefore, the role of medical training in Precision Medicine is also examined and analyzed. Specific examples are discussed in an integrated, precise, and data-driven manner to provide “smart” solutions.
  • Translating a value-based framework for resilient e-learning impact in post COVID-19 times: Research-based Evidence from Higher Education in Kuwait

    Afnan Alkhaldi; Sawsan Malik; Rashed Alhaimer; Abdullah Alshaheen; Lytras, Miltiadis; External Collaboration; NA; NA; NA; Computer Science; et al. (Elsevier, 2024-01-30)
    The covid-19 pandemic has changed people’s daily lives and ehaviors all across the world and has impacted practically every element of human existence. The introduction of remote education systems and the move toward online learning have had some of the most ignificant effects. The on-site operations of educational institutions, such as schools, colleges, and universities, have had to be suspended in order to stop the virus’ spread. In order to effectively disseminate instructional material and guarantee the unbroken progression of students’ academic endeavors, educators have been forced to look for novel approaches. The study used the Value-Based Adoption Model (VAM) as a conceptual framework to look into the factors that affected Kuwait’s e-learning outcomes in the midst of the covid-19 pandemic. 382 students at Kuwaiti universities and colleges were the source of quantitative data collection. The findings revealed that peer interaction emerged as the most influential factor in shaping utcomes within the educational context of Kuwait, while instructors and course design factors were not significant. Using the VAM, this study investigated the impact of several factors on students’ e-learning results during times of crisis. The research expands the existing knowledge base in the field on this subject and suggests developing a well-organized online learning crisis approach. The main contribution of this work is summarized on (i) An integrated framework for the quality of the e-learning experience in universities in post-covid-19 times and (ii) A resilient higher education institutional learning strategy model in post-covid-19 times. The findings of this paper can be generalizable to other Gulf Corporation Council (GCC) countries such as Kingdom of Saudi Arabia, Qatar, United Arab Emirates (UAE), Bahrain and Oman. This is due to the shared cultural traditions and values, along with similar educational systems among these nations.
  • Guest editorial: Gender, entrepreneurship and the digital divide: a global perspective

    Anna Visvizi; Lytras, Miltiadis; Chuman, Merwat; Sarirete, Akila; Kozłowski, Krzysztof; External Collaboration; Artificial Intelligence & Cyber Security Lab; 0; 0; Computer Science; et al. (Emerald Publishing Limited, 2023-10-11)
    The concepts and terms defining the thrust of this special issue, i.e. gender, entrepreneurship (Berger et al., 2021), the digital divide (Bowen and Morris, 2019; Millan et al., 2021; Satalkina et al., 2021), Global South (Simaan, 2020) and Global North, are very well established in the literature. Nevertheless, relatively little has been written about (1) the gendered dimension of the digital divide, (2) the digital divide and the gendered dimension of entrepreneurship (Elliott et al., 2021); and finally, (3) the specificity of these topics as they are in the Global South and Global North's peripheries (Ojediran and Anderson, 2020; Wood et al., 2021; Althalathini et al., 2020). Even if research on each of these individual domains exists, relatively little research on the intersection of these three areas exists (but cf. Visvizi et al., 2023, and earlier Kasusse, 2005; Alden, 2003). Notably, given the pace and the pervasive impact of digital transformation globally, and their diverse political, social and economic manifestations, it is necessary that the mechanisms underlying these interconnected issues and developments are identified and explored. This special issue sought to encourage this kind of conversation, always in context of the United Nation's (UN) Sustainable Development Goals (SDGs). Specifically, the key objective of this special issue was to examine, (1) how the onset of increasingly sophisticated information and communication technology (ICT) influences gender and entrepreneurship in the Global South and in the peripheries of the Global North; (2) what types of interpretive lens and explanatory potential are offered by the existing literature on the subject and (3) whether best practice-sharing and specific business and policy strategies might be helpful in alleviating negative implications of the global digital transformation.
  • Interpretable Detection of Malicious Behavior in Windows Portable Executables using Multi-Head 2D Transformers

    Khan, Sohail; Nauman, Mohammad; Department Collaboration; Artificial Intelligence & Cyber Security Lab; 0; 0; Computer Science; 0; Khan, Sohail (2023-11-17)
    Windows malware is becoming an increasingly pressing problem as the amount of malware continues to grow and more sensitive information is stored on systems. One of the major challenges in tackling this problem is the complexity of malware analysis, which requires expertise from human analysts. Recent developments in machine learning have led to the creation of deep models for malware detection. However, these models often lack transparency, making it difficult to understand the reasoning behind the model’s decisions, otherwise known as the black-box problem. To address these limitations, this paper presents a novel model for malware detection, utilizing vision transformers to analyze the opcode sequences of more than 350,000 Windows portable executable malware samples from real-world datasets. The model achieved a high accuracy of 0.9864, not only surpassing previous results but also providing valuable insights into the reasoning behind the classification. Our model is able to pinpoint specific instructions that lead to malicious behavior in malware samples, aiding human experts in their analysis and driving further advancements in the field. We report our findings and show how causality can be established between malicious code and actual classification by a deep learning model thus opening up this blackbox problem for deeper analysis.
  • An Efficient Encryption and Compression of Sensed IoTMedical Images Using Auto-Encoder.

    ElKafrawy, Passent; elsayed, Ayman; Torkey, Hanaa; Ahmed, Abdelmoty M.; Aboghazalah, Maie; External Collaboration; Virtual Reality Lab; 0; 0; Computer Science; et al. (EBSCO, 31 August )
    Healthcare systems nowadays depend on IoT sensors for sending data over the internet as a common practice. Encryption ofmedical images is very important to secure patient information. Encrypting these images consumes a lot of time onedge computing; therefore, theuse of anauto-encoder for compressionbefore encodingwill solve such a problem. In this paper, we use an auto-encoder to compress amedical image before encryption, and an encryption output (vector) is sent out over the network. On the other hand, a decoder was used to reproduce the original image back after the vector was received and decrypted. Two convolutional neural networks were conducted to evaluate our proposed approach:The first one is the auto-encoder, which is utilized to compress and encrypt the images, and the other assesses the classification accuracy of the image after decryption and decoding. Different hyperparameters of the encoder were tested, followed by the classification of the image to verify that no critical information was lost, to test the encryption and encoding resolution. In this approach, sixteen hyperparameter permutations are utilized, but this research discusses three main cases in detail. The first case shows that the combination of Mean Square Logarithmic Error (MSLE), ADAgrad, two layers for the auto-encoder, and ReLU had the best auto-encoder results with a Mean Absolute Error (MAE) = 0.221 after 50 epochs and 75% classification with the best result for the classification algorithm. The second case shows the reflection of auto-encoder results on the classification results which is a combination ofMean Square Error (MSE), RMSprop, three layers for the auto-encoder, and ReLU, which had the best classification accuracy of 65%, the auto-encoder gives MAE = 0.31 after 50 epochs. The third case is the worst, which is the combination of the hinge, RMSprop, three layers for the auto-encoder, and ReLU, providing accuracy of 20% and MAE = 0.485.
  • A Blockchain-Enabled IoT Logistics System for Efficient Tracking and Management of High-Price Shipments: A Resilient, Scalable and Sustainable Approach to Smart Cities

    Mohammed Balfaqih; Zain Balfagih; Miltiadis D Lytras; Khaled Mofawiz Alfawaz; Abdulrahman A Alshdadi; Eesa Alsolami; External Collaboration; Artificial Intelligence & Cyber Security Lab; 0; 0; et al. (MDPI, 2023-09-20)
    The concept of a smart city is aimed at enhancing the quality of life for urban residents, and logistic services are a crucial component of this effort. Despite this, the logistics industry has encountered issues due to the exponential growth of logistics volumes, as well as the complexity of processes and lack of transparency. Consequently, it is necessary to develop an efficient management system that offers traceability and condition monitoring capabilities to ensure the safe and high-quality delivery of goods. Moreover, it is crucial to guarantee the accuracy and dependability of distribution data. In this context, this paper proposes a blockchain-enabled IoT logistics system for the efficient tracking and management of high-price shipments. A smart contract based on blockchain technology has been designed for automatic approval and payment, with the aim of distributing shipping information exclusively among legitimate logistics parties. To ensure authentication, a zero-knowledge proof is used to conceal the blockchain address. Moreover, an intelligent parcel (iParcel) containing piezoresistive sensors is developed to pack delivered goods during the shipping process for violation detection such as severe falls or theft. The iParcels are automatically tracked and traced, and if a violation occurs, the contract is cancelled, and payment is refunded. The transaction fee per party is reasonable, particularly for high-price products that guarantee successful shipment.
  • Risk prediction model for cannabis use with artificial intelligence approach

    Hakkarainen, P; Karjalainen, K.; Subasi, A; External Collaboration; Artificial Intelligence & Cyber Security Lab; 0; 0; Computer Science; 0; Unlu, A. (Taylor and Francis, 01/08/2023)
    Background Identifying the most important predictors of substance use is crucial for developing effective prevention policies. Traditional statistical methods have some limitations in this regard. To address these limitations, the researchers utilized artificial intelligence (AI) methods to identify the top 10 most important predictors of cannabis use in Finland. Objective The objective of this study was to apply AI techniques to identify the key predictors of cannabis use in Finland. Specifically, the researchers aimed to determine the top 10 most important features related to cannabis use from a dataset consisting of 3229 observations and 313 questionnaire items, with 48 selected for preprocessing. Methods The researchers employed the recursive feature elimination (RFE) method as part of their AI analysis. This technique was used on 60 processed variables, following the application of missing data imputation, resampling, and scaling techniques. The RFE method allowed the researchers to narrow down the 60 variables to the top 10 most important features associated with cannabis use. Results The AI models developed using the selected features were able to predict cannabis use with a remarkable accuracy of 96% for the previous 12 months. The results of the study revealed that the social settings of individuals played the most significant role in predicting cannabis use in the context of Finland. Conclusions In conclusion, this study demonstrated the effectiveness of AI-based approaches in identifying the most critical predictors of cannabis use in Finland. The research highlighted that social settings had the highest impact on cannabis use in this setting. Moreover, the study showcased the potential of AI methods not only for identifying key risk indicators among various factors but also for optimizing the utilization of limited public resources when devising prevention strategies. These findings can be valuable for shaping targeted and efficient prevention policies to address cannabis use in Finland.
  • Promoting accuracy in low-magnification histopathology grading: With augmentation and multi-dilation model.

    Subasi, Abdulhamit; External Collaboration; Artificial Intelligence & Cyber Security Lab; 0; 0; Computer Science; 0; Gan, Zonghan (Elsevier, September )
    Advances in artificial intelligence have facilitated the automated grading of histopathology slides. Yet, the magnification of whole slide scanners (WSS) has restrained the accuracy of patch-based grading. In this work, we found that augmentation can significantly promote grading performance under this issue, even when the data volume is large (>140 K). With augmentation and a multi-dilation model, the CovXNet, we yielded a Balanced Accuracy of 92.13%, which is the current highest for the Breast Histopathology Dataset (40X magnification) also the first time both sensitivity and specificity >90%. However, in this focused grading task, augmentation only improves models with high invariance (the CovXNet and BCA-CNN). Pre-trained ResNet has lower invariance in this task, but fine-tuning can significantly improve both accuracy and invariance. For the CropNet attention model, adapting with max pooling but not augmentation offers promotions. Additionally, this work also found two types of common errors in high-starred codes, when using random.shuffle for data-label composited array, or the integrated shuffle function of ImageDataGenerator, which fake a higher accuracy by masking class 0 as class 1. Using Sklearn.shuffle instead is safer. All codes are available on our GitHub.
  • A Blockchain-Enabled IoT Logistics System for Efficient Tracking and Management of High-Price Shipments: A Resilient, Scalable and Sustainable Approach to Smart Cities

    Balfaqih, Mohammed; Balfagih, Zain; Lytras, Miltiadis; Alfawaz, Khaled; Alshdadi, Abdulrahman; Alsolami, Eesa; External Collaboration; Artificial Intelligence & Cyber Security Lab; 0; 0; et al. (MDPI, 2023-09-20)
    The concept of a smart city is aimed at enhancing the quality of life for urban residents, and logistic services are a crucial component of this effort. Despite this, the logistics industry has encountered issues due to the exponential growth of logistics volumes, as well as the complexity of processes and lack of transparency. Consequently, it is necessary to develop an efficient management system that offers traceability and condition monitoring capabilities to ensure the safe and high-quality delivery of goods. Moreover, it is crucial to guarantee the accuracy and dependability of distribution data. In this context, this paper proposes a blockchain-enabled IoT logistics system for the efficient tracking and management of high-price shipments. A smart contract based on blockchain technology has been designed for automatic approval and payment, with the aim of distributing shipping information exclusively among legitimate logistics parties. To ensure authentication, a zero-knowledge proof is used to conceal the blockchain address. Moreover, an intelligent parcel (iParcel) containing piezoresistive sensors is developed to pack delivered goods during the shipping process for violation detection such as severe falls or theft. The iParcels are automatically tracked and traced, and if a violation occurs, the contract is cancelled, and payment is refunded. The transaction fee per party is reasonable, particularly for high-price products that guarantee successful shipment.

View more