This article explores the potential for addressing issues related to improving the sustainability and safety of a city's transportation system. The study demonstrates the need for a comprehensive approach that considers not only the direct but also the indirect effects of potential solutions, selecting those that enhance the overall sustainability and safety of the system. The article presents a conceptual model of such a system and illustrates implementation methods using a large city as an example. Furthermore, it demonstrates how simulation modeling can address urban transportation network issues and quantifies the impact of implementing the proposed solutions.

Due to growing environmental concerns in relation to global warming and pollutants, it’s become very critical to study about sources which are contributing towards it. Once such source is emissions from the industrial equipment’s like Gas Turbine (GT). GTs are used all over the world across various industries for power generation or mechanical drive operations. Some of these industries belongs to Energy, Utilities, Refineries, Chemical & Fertilizer sectors. In this study, efforts were be made to compare all commonly used ML regression models such as MLR, DTR, RFT, Adaboost Regressor, GBR, XGBoost Regressor using same dataset for emission prediction (CO & NOx). Model’s evaluation metrices were analyzed to predict the best model along with number of significant features. Evaluation metric included MAE, RAE, MAPE & RMSE. In addition, various features were analyzed for patterns using univariate & multivariate analysis graphical tools. This study can contribute towards new PEMS (Prediction Emission Monitoring System) designing using ML Regression Models which will help industrial companies in boosting their operation efficiencies, minimization of emissions (like CO & NOx) and making financial gains due to its cost effectiveness over traditional CEMS (Continuous Emission Monitoring System). Overall, these new PEMS will keep monitoring and optimizing industrial equipment’s operations for emissions control during their lifecycles.

Plant diseases pose a significant threat to global food production and require early identification to ensure global food security and safeguard the economic interests of agriculture stakeholders. Practical implementation in agriculture requires fast, lightweight and accurate models for edge devices. This research presents a tomato leaf disease classification approach using a fast and lightweight pre-trained CNN architecture designed for efficiency. The method incorporates resampling techniques to address the data class imbalance, while pre-processing steps such as resizing, and augmentation enhance model performance. The transfer learning-based NASNetMobile model is trained and evaluated using the plant village dataset, containing nine disease classes and a healthy class. Performance assessment focuses on lightweight characteristics and classification accuracy of the model. Results demonstrate the effectiveness of the approach in balancing lightweight characteristics and accuracy, offering the potential for extending disease classification to other plant species. This research contributes to the development of low-end devices for easy disease identification, benefiting global food security and agricultural stakeholders’ economic well-being.

Facial masks have become essential in our daily lives, especially at airports, hospitals, and other public places. Post-pandemic, we observe people wearing masks to protect themselves from dust, viral infections, and pollution. While face recognition is a well-studied problem, Masked Face Recognition (MFR) is challenging due to the prominent part of the face being missing or occluded by the mask. In the literature, fine-tuned deep-learning models for MFR have been proposed. Many individual deep-learning networks have been built and fine-tuned to work with MFR datasets. However, they struggle to identify real-time masked faces in uncontrolled environments. This paper introduces an approach for real-time MFR detection based on Few-Shot Learning (FSL) using the Labeled Faces in the Wild (LFW) dataset. Our model excels at identifying both unmasked and masked faces in real-time scenarios. The current practices in normal face recognition or masked face recognition datasets make it challenging to register new masked faces and identify them in real-time. To address this problem, we propose a customized FSL dataset that combines LFW and MLFW (Masked LFW) and includes an augmented dataset featuring both face-to-mask and mask-to-face transformations. In real-time face registration and identification scenarios, it is crucial to train our model using only a few shots. The few-shot data used for registering faces are augmented using a pre-trained face completion model. This augmented data is then processed by existing deep-face models to generate embeddings, which are further trained using an SVM classifier for masked face identification. In comparison to baseline models, our proposed approach for masked face identification has significantly improved accuracy by 10%, achieving an accuracy rate of 87.52% with FaceNet512 compared to previous methods. We also provide real-time performance readings at various stages of our approach and report promising inference times for MFR using the final model, which is as low as 0.024 seconds.

COVID-19 crisis has led to an outburst of information that needs to be organized, validated, and made available to the seekers. Despite the rapid growth and success of BERT models in the last 3 years, COVID QA is a difficult task due to the lack of applicable datasets and a relevant language representation. Therefore, this study proposes a transformer-based Question Answering (QA) model for COVID-19 questions from the biomedical domain. Further, explored several datasets, and models required for question type prediction, no-answer prediction, and answer extraction and transfer learning strategies. It has been demonstrated that the exact match score can be significantly improved with limited amounts of training data from the biomedical domain. Finally, the findings of the study have been summarized as Factoid QA Finetuning Framework (FQFF), which can provide initial direction for domain-specific QA tasks with a limited amount of data.

This research proposes a computer vision-based solutions to identify whether a patient is covid19/normal/Pneumonia infected with comparable or better state-of-the-art accuracy. Proposed solution is based on deep learning technique CNN (Convolutional Neural networks) with multiple approaches to cover all open issues. First approach is based on CNN models based on pre-trained models; second approach is to create CNN model from scratch. Experimentation and evaluation of multiple approaches helps in covering all open points and gaps left unattended in related work performed to solve this problem. Based on the experimentation results of both the approaches and study of related work done by other researchers, Both the approaches are equally effective can be recommended for multi-class classification of lung disease.

The devastating earthquakes that struck Syria and Turkey On 6 February 2023 caused severe damage bringing life to a standstill. The seriousness of the situation led the President of Turkey to instantly announce a three-month-long state of emergency across the nation. The United Nations Office of the Coordination of Humanitarian Affairs declared that severe weather conditions accompanied by rising temperatures have made relief work difficult and lengthy. The two countries had no option but to request international assistance. Social Media, especially Twitter, was vastly used by persons from various walks of life for ground reporting, seeking the attention of government and private agencies for aid and support, outpouring emotions, drawing international attention, and more. In this regard, exploration of data from such posts can reveal people’s sentiments and bring to the forefront vital information for the concerned agencies to take action and respond accordingly. The prediction of topics will facilitate the bifurcation of information and further boost the restoration process through immediate task allocation to concerned relief societies. The application of the best-performing machine learning algorithm to the data set will bring out significant traits from the data to have a finer semantic understanding and facilitate the development of a high-performing topic-based prediction model.

However, energy forecasting is still a challenging task due to the many unpredictable factors that can impact energy consumption and production, such as changes in weather patterns, economic conditions, and energy policies. Therefore, energy forecasts should be continuously updated and refined as new information becomes available. The purpose of this research is to present a high-level, machine learning-centric viewpoint on data-driven energy forecasting. Challenges and constraints of data-driven energy forecasting will be discussed, along with the many machine learning methods and methodologies that can be implemented. The paper compares the performance of various deep learning and time series analysis techniques such as LSTM, RNN, ARIMA and SARIMA in energy forecasting. This research can provide a more comprehensive understanding of the effectiveness of different models in energy forecasting, which can have significant implications for energy management, policymaking, and infrastructure development.

The deadly virus COVID-19 has heavily impacted all countries and brought a dramatic loss of human life. It is an unprecedented scenario and poses an extreme challenge to the healthcare sector. The disruption to society and the economy is devastating, causing millions of people to live in poverty. Most citizens live in exceptional hardship and are exposed to the contagious virus while being vulnerable due to the inaccessibility of quality healthcare services. This study introduces ubiquitous computing as a state-of-the-art method to mitigate the spread of COVID-19 and spare more ICU beds for those truly needed. Ubiquitous computing offers a great solution with the concept of being accessible anywhere and anytime. As COVID-19 is highly complicated and unpredictable, people infected with COVID-19 may be unaware and still live on with their life. This resulted in the spread of COVID-19 being uncontrollable. Therefore, it is essential to identify the COVID-19 infection early, not only because of the mitigation of spread but also for optimal treatment. This way, the concept of wearable sensors to collect health information and use it as an input to feed into machine learning to determine COVID-19 infection or COVID-19 status monitoring is introduced in this study.

Gesture recognition is a topic in computer science and language technology with the goal of interpreting human gestures via mathematical algorithms. It is a subdiscipline of computer vision. In this paper, we describe some of Gesture recognition techniques such as Vision based gesture recognition and Graph based gesture recognition. Also, we explore these techniques with previous studies.

Gesture recognition is a computing process that attempts to recognize and interpret human gestures through the use of mathematical algorithms. In this paper, we describe Point Based Gesture Recognition and Point Clouds nearest neighbors and sampling. Also, we explore these techniques with previous studies.

Breast Cancer is the second most leading cause of death among women. The early detection of the disease increases the chances of survival of the patient. Therefore, there is always a need for techniques that can accurately predict the presence of cancer. Data Mining is one such powerful technique that can assist clinicians to effectively use the data for timely prediction of the disease. In the medical domain, data is usually imbalanced with unequal distribution of the positive and negative classes. Imbalanced datasets introduce a bias in the model and can thus reduce the accuracy of the minority class predictions. In the case of cancer detection, the mammographic data is highly imbalanced, and predicting the positive (minority) class is of the utmost importance. To achieve this, different models using various class balancing techniques are built and evaluated. The experiments show that the performance of the weighted approach and the undersampling technique is better than oversampling and hybrid techniques. The best performing classifiers are the weighted XGBoost model and Stacking ensemble with the average AUC of 0.78 and 0.76 respectively.

The education domain is growing at an exponential rate, maturing with the introduction of innovative and improved offerings to the learning community aided by various Education Data Mining techniques (EDM). Significant amount of research is being carried out around EDM, using various dimensions like student's performance, dropout rates, individual cognitive capabilities, teacher's, administrators' performance, course delivery, and content authoring. Several data analytics techniques are adopted to gain knowledge from educational data. However, the prediction models generated are complex and not interpretable in understanding why and how a prediction has arrived at from the produced results. The interpretability of the model becomes increasingly important when dealing with large datasets and complex models. The derived analytics and the models will be acceptable and trusted if they are interpretable. It helps the educational organizations to advance their learning processes in boosting students' performance and reduction in the student dropout rate. This study reviews EDM focusing on the factors influencing student's predictions, various algorithms used, and identified the gaps. The study also gives an insight into how the "black-box" decisions of the prediction model are made, the role of various eXplainable AI (XAI) techniques in making the model results interpretable, and their contribution to producing explainable results.

Information security provides a set of mechanisms to be implemented in the organisation to protect the disclosure of data to the unauthorised person. Access control is the primary security component that allows the user to authorise the consumption of resources and data based on the predefined permissions. However, the access rules are static in nature, which does not adapt to the dynamic environment includes but not limited to healthcare, cloud computing, IoT, National Security and Intelligence Arena and multi-centric system. There is a need for an additional countermeasure in access decision that can adapt to those working conditions to assess the threats and to ensure privacy and security are maintained. Risk analysis is an act of measuring the threats to the system through various means such as, analysing the user behaviour, evaluating the user trust, and security policies. It is a modular component that can be integrated into the existing access control to predict the risk. This study presents the different techniques and approaches applied for risk analysis in access control. Based on the insights gained, this paper formulates the taxonomy of risk analysis and properties that will allow researchers to focus on areas that need to be improved and new features that could be beneficial to stakeholders.