We present a novel theory of the microstructure of deep neural networks. Using a theoretical framework called \textit{statistical teleodynamics}, which is a conceptual synthesis of statistical thermodynamics and potential game theory, we predict that all highly connected layers of deep neural networks have a universal microstructure of connection strengths that is distributed lognormally ($LN (\mu, \sigma$)). Furthermore, under ideal conditions, the theory predicts that $\mu$ and $\sigma$ are the same for all layers in all networks. This is shown to be the result of an arbitrage equilibrium where all connections compete and contribute the same effective utility towards the minimization of the overall loss function. These surprising predictions are shown to be supported by empirical data from six large-scale deep neural networks in real life. We also discuss how these results can be exploited to reduce the amount of data, time, and computational resources needed to train large deep neural networks.

[ACCEPTED FOR PUBLICATION IN THE DAIS JOURNAL OF SCIENCE 2023] Intercity travel optimization has always been a goal of every government. Such an optimization would allow efficient and economical development. Optimizing travel would facilitate transfer of ideas and resources and promote economic growth. As such, travel can be optimized on two fronts: cost of arranging for that travel and the ease of travel i.e., the length/time of travel for a person to go from any one city to any other. This study proposes a methodology of optimizing travel pathways in any circumstance through genetic algorithm metaheuristics and network theory representations. This study demonstrates the same to optimize inter-city travel between the 25 cities with the highest population in India.

A computer program was built to model a random walk that represents utilization of a limited set of items by a group of servers for  their customers. An example is the use of ventelators (items) for very sick patients (customers)  by hospitals (servers). Some hospitals amy run out of ventilators while others may have extras if a common pool of ventilators is not employed. The simulation showed a 5% improvement in one of the cases studied. A theoretical model was developed that displayed very accurate prediction of the simulation results.

The exploration and categorization of essential and synthetic lethality genes hold significant importance in seeking effective and targeted therapies for diverse ailments. This endeavor hinges upon genetic minimal cut sets (gMCSs), which also find utility in metabolic engineering. There have been various methods suggested for calculating gMCSs. Still, with the emergence of numerous new models and their growing intricacy, it has become vital to introduce new algorithms in this field. This paper presents a new algorithmic approach for computing gMCSs, which utilizes linear programming techniques to improve temporal efficiency. The key concept of the method is to use a k-representative subset to replace the target set with a smaller one.

A lightweight cylindrical Frequency Selective Surface (FSS) using high-pass inductive mesh grid is proposed as a radiofrequency (RF) shield for solenoid coils at 2.84 MHz working in a portable MRI system with a transversal B0-field averaged at 67 mT. Using a solenoid coil of 60 mm diameter as an example, the proposed FSS shield shows good shielding effectiveness at different wall-to-wall distances between the coil and the shield (5-25 mm). It also shows less compromise in B1-sensitivity when compared to the coil with a copper shield. The effects of the FSS shield were examined numerically and validated experimentally. When the wall-to-wall distance is 5 mm, the solenoid coil with the proposed FSS shield is compact, lightweight, and shows B1-sensitivity of more than 50% higher than the copper counterparts with comparable shieding effectiveness. Compared to an unshielded coil, it shows a wider bandwidth which benefits the excitation of an MRI system that has less homogeneous fields and a smaller inductance which indicates less trapping of energy and thus can enable different pulse sequence programming by taking this approach. This study shows that an FSS shield with inductive mesh grids around a solenoid coil is a promising approach for compact shielding without a big shielding box. It will contribute to the compactness of a portable MRI system.

Road traffic crashes in urban areas pose significant safety and efficiency challenges. This paper presents an approach, merging case studies and literature reviews, to mitigate urban road traffic crashes and enhance safety. The study commences with an extensive literature review, delving into existing research on road traffic crashes. This review establishes the groundwork for comprehending urban road safety solutions, insights into ongoing research trends, and guiding case study selection. Three case studies parallel analogous cities where Intelligent Transportation Systems and other smart city initiatives targeting road safety succeeded. Each case dissects safety interventions, accomplishments, and lessons learned. Scrutinizing these studies aims to unveil trends underpinning a holistic road safety strategy. From literature and case studies, a comprehensive road safety strategy emerges. It leverages advanced technology, data insights, and proactive measures to enhance road safety, curtail accidents, and optimize urban traffic. This research yields an all-encompassing road safety strategy for policymakers, planners, and safety practitioners to elevate urban road safety. It promotes secure, efficient, technologically advanced urban areas, endorsing sustainable transportation, elevating residentsâ\euro™ lives, and positioning cities as urban mobility leaders. Furthermore, this study contributes to the academic literature on road safety and urban transportation. Through case studies and literature reviews, it supplements knowledge about effective road traffic crash mitigation strategies. These insights have the potential to guide future research, advancing road safety and urban transportation domains.

Accurate trajectory prediction is vital for various applications, including autonomous vehicles. However, the complexity and limited transparency of many prediction algorithms often result in black-box models, making it challenging to understand their limitations and anticipate potential failures. This further raises potential risks for systems based on these prediction models. This study introduces the Performance Monitor for Black-Box Trajectory Prediction Model (PMTM) to address this challenge. The PMTM estimates the performance of black-box trajectory prediction models online, enabling informed decision-making. The study explores various methods’ applicability to the PMTM, including anomaly detection, machine learning, deep learning, and ensemble, with specific monitors designed for each method to provide real-time output representing prediction performance. Comprehensive experiments validate the PMTM’s effectiveness, comparing different monitoring methods. Results show that the PMTM effectively achieves promising monitoring performance, particularly excelling in deep learning-based monitoring. It achieves improvement scores of 0.81 and 0.79 for average prediction error and final prediction error monitoring, respectively, outperforming previous white-box and gray-box methods. Furthermore, the PMTM’s applicability is validated on different datasets and prediction models, while ablation studies confirm the effectiveness of the proposed mechanism. Hybrid prediction experiments further demonstrate the value of the proposed monitor from an application perspective.

Background: Although early rehabilitation is important following a stroke, severely affected patients have limited options for intensive rehabilitation as they are often bedridden. To create a system for early rehabilitation of lower extremities in severely affected patients, we have combined the robotic manipulator ROBERT® and EMG-triggered FES and developed a novel user-driven Assist- As-Needed (AAN) control approach. The method is based on a state machine that can detect user movement capability and provide different levels of assistance, as required by the patient (no support, FES only, and simultaneous FES and mechanical support). Methods: To technically validate the system, we tested 10 able-bodied participants who were instructed to perform specific behaviors to trigger the desired system states while conducting knee extension and ankle dorsal flexion exercise. In addition, the system was tested on two stroke patients to establish the clinical feasibility. Results: The technical validation showed that the state machine correctly detected the participants’ behavior and activated the target AAN state in more than 96% of the exercise repetitions. The clinical feasibility test showed that the system successfully recognized the patients’ movement capacity and activated assistive states according to their needs, providing the minimal level of support required to perform the exercise successfully. Conclusions: The system was technically validated and preliminarily proven clinically feasible. The present study shows that the novel system can be used to deliver exercises with a high number of repetitions while engaging the participants’ residual capabilities through an effective AAN strategy.

A document by Wazib Ansar . Click on the document to view its contents.

This paper presents a novel loss function referred to as hybrid ordinary-Welsch (HOW) and a new sparsity-inducing regularizer associated with HOW. We theoretically show that the regularizer is quasiconvex and that the corresponding Moreau envelope is convex. Moreover, the closed-form solution to its Moreau envelope, namely, the proximity operator, is derived. Compared with nonconvex regularizers like the lp-norm with 0

The monkeypox virus poses a novel public health risk that might quickly escalate into a worldwide epidemic. Machine learning (ML) has recently shown much promise in diagnosing diseases like cancer, finding tumor cells, and finding COVID-19 patients. In this study, we have created a dataset based on the data both collected and published by Global Health, and used by the World Health Organization (WHO). Being entirely textual, this dataset shows the relationship between the symptoms and the monkeypox disease. The data has been analyzed, using gradient boosting methods such as Extreme Gradient Boosting (XGBoost), CatBoost, and LightGBM along with other standard machine learning methods such as Support Vector Machine (SVM) and Random Forest. All these methods have been compared. The research aims at providing an ML model based on symptoms to diagnose monkeypox. Previous studies have only examined disease diagnosis-using images. The best performance has belonged to XGBoost, with an accuracy of 1.0 in reviews. To check the model’s flexibility, k-fold cross-validation is used, reaching an average accuracy of 0.9 in 5 different split of test set. In addition, Shapley Additive Explanations (SHAP) helps examining and explaining the output of the XGBoost model.

The Null result of the Michelson & Marley Experiment is explained by the invariant wavelength of light while propagating or transiting from one inertial reference frame to another. It is shown that under classical kinematics wavelength remains constant as observed by two reference frames. Thus the explanation for the Null result of the M&M experiment is the simpleset under classical kinematics.

In the field of multimodal semantic understanding, tackling inherent uncertainties is essential for mitigating ambiguous interpretations across multiple targets. This paper introduces the Probability Distribution Encoder (PDE), a versatile, plug-and-play module that utilizes sequence-level and feature-level interactions to model these uncertainties as probabilistic distributions. We demonstrate its adaptability by seamlessly integrating PDE into established frameworks, culminating in models like SWINPDE. Compared to previous methods, our probabilistic approach substantially enriches multimodal semantic understanding. We incorporate this uncertainty modeling into prevalent pre-training architectures and propose specialized pre-training tasks: Distribution-based Vision-Language Contrastive Learning (D-VLC), Distribution-based Masked Language Modeling (D-MLM), and Distribution-based Image-Text Matching (D-ITM). Empirical tests show that our models achieve state-of-the-art (SoTA) results in a range of downstream tasks, including image-text retrieval, visual question answering, visual reasoning, visual entailment and video captioning. Furthermore, our qualitative results reveal several superior properties conferred by this uncertainty modeling, such as enhanced semantic expressiveness compared to point representations, and the ability to generate diverse yet accurate predictions.

M-estmators including the Welsch and Cauchy have been widely adopted for robustness against outliers, but they also down-weigh the uncontaminated data. To address this issue, we devise a framework to generate a class of nonconvex functions which only down-weigh outlier-corrupted observations. Our framework is then applied to the Welsch, Cauchy and lp-norm functions to produce the corresponding robust loss functions. Targeting on the application of robust matrix completion, efficient algorithms based on these functions are developed and their convergence is analyzed. Finally, extensive numerical results demonstrate that the proposed methods are superior to the competitors in terms of recovery accuracy and runtime.

In the intricate and ever-evolving realm of cybersecurity, the safeguarding of digital assets and information becomes paramount, especially considering the sophistication of contemporary cyber threats. This research embarks on a novel exploration of integrating Blockchain, Java, and Artificial Intelligence (AI) to formulate a robust cybersecurity framework, metaphorically likened to the deductive prowess of the fictional detective, Sherlock Holmes. Blockchain technology is heralded for its immutable and decentralized characteristics, offering a fortification against data tampering. Java, with its platform-independent and object-oriented programming capabilities, provides a versatile and secure environment for developing blockchain applications. Concurrently, AI introduces an adaptive and predictive element, enabling the system to anticipate, identify, and mitigate potential cyber threats through pattern recognition and anomaly detection. Through a series of case studies and empirical analyses, this research elucidates the efficacy of the proposed cybersecurity framework in thwarting various cyber-attack vectors, ensuring data integrity, and maintaining user privacy. Furthermore, it navigates through the ethical and legal conundrums associated with deploying intelligent and autonomous cybersecurity systems. The findings underscore the potential of amalgamating Blockchain, Java, and AI in enhancing cybersecurity, providing insights into overcoming associated challenges, and paving the way for future research in deploying intelligent, secure, and ethical cybersecurity frameworks.

The research employs the Uses and Gratification model to identify three key impacts of COVID-19 related misinformation on social media users: (1) an “infodemic” of health-related misinformation; (2) an increase in social media pastime; and (3) an increase in social, political, and religious conflicts. Several actions are proposed to limit the spread of fake news, involving entities such as government and health agencies, as well as social media companies

Continuous monitoring and prediction of Length of Stay (LoS) for critically ill patients admitted to Intensive Care Units (ICU) can help anticipate on their needs, thereby reducing risks of adverse events occurring, efficiently allocate resources and improve care. In this work, four different sequence-to-sequence deep learning models (Long-Short Term Memory, Gated Recurrent Units, Temporal Convolution Network and the Transformer network) were implemented to predict hourly remaining LoS, using 271 input features extracted from 20,489 ICU stays. Input sequences for model training were created using two approaches: (i) by stacking data one hour at a time and (ii) by using a sliding window of length $l$ shifted one hour at a time. To enhance model explainability, both the models’ latent space visualization and feature importance ranking were used. Compared against baseline models, the sequential models reported the best performance, reducing the prediction error by almost 3 days. Moreover, the sliding window approach showed improvement both in terms of performance and training time whilst requiring less resources. The latent space visualization of each model using t-SNE revealed that, with only 6 hours of data, the models are able to learn early clinical patterns that separate short ($0<\text{LoS}\leq3$ days) from medium ($3<\text{LoS}<7$ days) and long ($7\leq\text{LoS}<+\infty$ days) stays, and, the models are able to capture existing correlations between long LoS and risk of mortality, the latter not used during modeling. The benefit of this approach in a true hospital setting, besides LoS prediction, is that early clustering enables the early identification of long stays with shown existing risk of death, to whom prompt allocation of focus and adequate resources can be delivered.

Often overlooked as a component of urban development, accessibility infrastructure is undeniably crucial in daily life. Accessibility ramps are one of the most common types of accessibility infrastructure, and serve to benefit not only people with mobile impairments but also able-bodied third parties. While the necessity of accessibility ramps is acknowledged, actual implementation fails in light of the limits of manpower required for the design stage. In response, we present an algorithm capable of the automatic generation of a feasible accessibility ramp based on a 3D model of the relevant environment. Through the manual specification of initial and terminal points within a 3D model, the algorithm uses AI search algorithms to determine the optimal pathway connecting these points. Essential components in devising a wheelchair-accessible ramp are encoded within the process, as evaluated by the algorithm, including but not limited to elevation differentials, spatial constraints, and gradient specifications. From this, the algorithm then generates the pathway to be expanded into a full-scale, usable model of a ramp, which then can be easily exported and transformed through inter-software exchanges. Though some human input is still required following the generation stage, the minimising of human resources provides significant boosts of efficiency in the design process thus lowering the threshold for the incorporation of accessibility features in future urban design.