Search

Background /purpose: Establishing an accurate postmortem interval (PMI) is exceptionally crucial in forensic investigation. Artificial intelligence (AI) and Machine learning (ML) models are widely employed in forensic practice. ML is a part of AI, both terms are highly associated and sometimes used interchangeably. This systematic review aims to evaluate the application and performance of AI technology for the prediction of PMI., Methods: Systematic literature search across different electronic databases using PubMed/Google Scholar/EMBASE/Scopus/CINAHL/Web of Science/Cochrane library was conducted from inception to 3 December 2021 for preclinical and clinical studies reported ML models for PMI estimation., Results: We identified 18 studies (12 preclinical and 06 clinical) that met the inclusion criteria in the qualitative analysis. Most of the studies employed supervised learning (N = 15), and others employed unsupervised learning (N = 3). Due to the heterogeneity of the samples, quantitative analysis was not performed., Conclusion: In this systematic review, we discussed the performance of AI-based automated systems in PMI estimation. ML models have demonstrated accuracy and precision and the ability to overcome human errors and bias. However, the research is limited, conducted in primarily small, selected human populations. In addition, we suggest further research in larger population-based studies is needed to fully understand the extent of integrated ML models., Competing Interests: Declaration of Competing Interest None., (Copyright © 2022 Elsevier B.V. All rights reserved.)

The recognition of individual activity has proven its importance in many application areas. Even after the pandemic crisis worldwide, the remote monitoring of human actions and their activities has increased a lot. In such situations, the smart sensors are utilized to capture the details remotely, share, and process intelligently using the computer-assisted machine. This paper introduces the new variant of the machine learning classifier, named Grid Search Based Nonlinear Multi-task Least Squares Twin Support Vector Machine (GMLT-SVM), to recognize human actions using sensory data. The proposed approach uses the grid search for hyperparameter tuning with a Nonlinear kernel for Multi-task classification in a twin support vector machine. This not only increases the efficacy of the model but also reduces the computation time for finding the optimal values of the hyper-parameters. The proposed approach when experimented with the sensory data of Human Action Recognition (HAR) gives an accuracy of 96.8%. The comparative analysis of the proposed method with existing other ML classifiers gives an increase in the accuracy of 94.2±2.6%. The experimental results suggest that the gained results are statistically significant enough to recognize human actions with sensory data. [ABSTRACT FROM AUTHOR]

In recent days, people must deal with stress brought on by the demands of modern living, which constantly presents new obstacles. Stress, a state of mental tension triggered by challenging circumstances, has become a global risk factor impacting individual well-being. Understanding variations in stress resilience is crucial for tailoring treatment strategies. Previous studies have explored stress prediction using measures like electroencephalography (EEG), blood pressure (BP), heart rate (HR), and interventions such as Kriya Yoga and mindfulness meditation. The experimentation is done on the data collected from people who practice heartfulness meditation regularly. The research employs machine learning (ML) algorithms alongside physiological parameters such as EEG, BP, HR, and psychological parameters, perceived stress scale (PSS), to precisely classify, measure, and predict stress levels. The investigations are done using K-nearest neighbor (KNN), random forest (RF), and kernel-support vector machine (k-SVM). An accuracy of 98.27% accuracy was achieved with the RF algorithm in classifying stressed and non-stressed individuals. [ABSTRACT FROM AUTHOR]

The surging popularity of generative adversarial networks (GANs) has ignited a wave of innovation in the realm of computer vision, a highly explored subfield of deep learning. GANs are revolutionizing the area of machine learning because they use a game-based training technique. This is in contrast to traditional approaches to machine learning, which center on feature learning and picture production. Several subfields of computer vision have seen tremendous progress thanks to the integration of numerous processing approaches, including image processing, dynamic processing, text, audio, and video processing, as well as generalized generative adversarial networks (GANs). Nevertheless, despite the fact that GANs have made great progress, they still offer promise that has not been fully realized and space for additional development. GANs have a wide range of applications within computer vision, including data augmentation, displacement recording, dynamic modeling, and image processing. This article digs into recent advances made by GAN researchers working in the realm of AI-based security and defense and discusses their accomplishments. In particular, we investigate how well image optimization, image processing, and image stabilization are incorporated into GAN-driven picture training. We want to achieve our goal of providing a complete overview of the present status of GAN research by carefully evaluating research articles that have been subjected to peer review. [ABSTRACT FROM AUTHOR]

This paper aims to establish a panoramic foundation for investigating the impact of sunlight on a recently formulated water-based tri-nano hybrid Sutterby liquid (TNHF) under provided magnetic field, directed through photovoltaic solar panels by exploiting the knacks of machine intelligent computing paradigm. The comparative performance of hybrid and tri-nano fluidic system with water as a base fluid is exhaustively analyzed and discussed. This research is unique, as it has a distinguished figurative comparison analysis between two different types of nanofluidic materials, which helps to choose the best one for use in agrivoltaics systems, and replace the materials already in use to enhance the efficiency and performance coefficients. Furthermore, the description of the composition scheme makes this research more feasible and applicable. A numerical dataset of nonlinear mathematical model is generated by employing finite difference scheme in the recently introduced Python bvp-solver algorithm, then it is embedded into artificial intelligence (AI)-based Levenberg Marquardt neural network algorithm (LMNNA). A significant outcome of the research indicates that the integration TNHF results in a notably faster enhancement of heat transfer rate and temperature framework as compared to traditional hybrid fluid. It is observed that introducing three distinct nanomaterials of specific thermophysical characteristics enhances the thermal exchanging profile and faces an obvious flow rate dissipation in solar plate channels. The standard numerical and AI-generated results are documented to portray the stability, accuracy and efficacy of scheme in terms of iterative learning curves on MSE, error analysis, histograms and regression statistics. Additional perquisites of the methodology include cost effectiveness, time-saving ability, robustness, stability and its extendibility. [ABSTRACT FROM AUTHOR]

Agriculture is one of the prime sources which determines a nation's prosperity and economic growth. But most of the farmers find it difficult to survive due to the following reasons like lack of enough labors, inappropriate and traditional methods are followed to monitor agricultural fields and plant growth etc. As technology has evolved, it is possible to automate the various phases of farming activities. There are many standard sensors available to remotely monitor the soil moisture content, pH content, fire detection in the field, intruder detection etc. All these sensors sense the information and is fed to a central coordinator which will process the data and initiate necessary actions. All these are done over internet and thus the concept can be called as Smart farming using Internet of Things. It is also possible to analyze the image of plant leaf to identify the various plant dis-eases with the help of Machine Vision. Our model thus reduces the farmer's overhead using IoT and computational intelligence. [ABSTRACT FROM AUTHOR]

Multimodal biometric systems have been widely used to achieve high recognition accuracy. This paper presents a comparison between three multimodal biometric systems used to authenticate human by fusion of palm dorsal hand and finger veins which used alternately. We developed an image analysis technique to extract region of interest (ROI) from palm dorsal hand and finger veins image. After extracting ROI we design a sequence of preprocessing steps to improve palm, dorsal hand and finger veins images using Homomorphic, Median filter, Wiener filter and Contrast Limited Adaptive Histogram Equalization (CLAHE) to enhance vein image. Our smart systems are based on the following algorithms, namely; principal component analysis (PCA) algorithm for feature extraction and k-Nearest Neighbors (KNN) classifier for matching operation. The database chosen was the CASIA Multi-Spectral Palm print Image Database V1.0,Bosphorus Hand Vein Database and the Shandong University Machine Learning and Applications - Homologous Multi-modal Traits (SDUMLA-HMT). The achieved results for the fusion biometric traits were Correct Recognition Rate (CRR) finger and dorsal hand veins is 96.8%., palm and dorsal hand veins is 97.6%, palm and finger veins is 97.9% [ABSTRACT FROM AUTHOR]

Multimodal biometric systems roughly used to achieve extreme recognition accuracy. This paper reports a novel multimodal biometric system employing intelligent technique to authenticate human by fusion of dorsal hand, palm and finger veins pattern. We improved an image analysis technique to separate region of interest (ROI) from dorsal hand, palm and finger veins image. After separating ROI we construct a sequence of preprocessing steps to enhance dorsal hand, palm and finger veins images using Median filter, Wiener filter, Contrast Limited Adaptive Histogram Equalization (CLAHE) and Homomorphic filter to improve vein image. Our intelligent technique is based on the following intelligent algorithms, namely; principal component analysis (PCA) algorithm for feature extraction and k-Nearest Neighbors (K-NN) classifier for matching operation. The database selected was Bosphorus Hand Vein Database, CASIA Multi-Spectral Palmprint Image Database V1.0 (CASIA database) and the Shandong University Machine Learning and Applications - Homologous Multi-modal Traits (SDUMLA-HMT). The accomplished result for the fusion of three biometric traits was Correct Recognition Rate (CRR) is 99.21% with False Reject Rate (FRR) 0.04%. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Remote Sensing is the property of Editorial Office of Journal of Remote Sensing & Science Publishing Co. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

With the popularity of computers and the rapid development of various application platforms, the explosive growth of data poses a huge challenge to data analysis and storage. For large-scale image analysis applications, the time delay in storing read data becomes an important issue that constrains this application. The semantic information asymmetry between image application and storage is the root cause of this problem. In view of content semantic analysis, in recent years, intelligent computing has become the main research direction. Among them, machine learning has become a research hot spot because of its offline learning and online generation characteristics. For the semantics of image content, machine learning can complete tasks such as content semantic association, classification, annotation and hash mapping, and provide algorithm support for applying image semantics and improving semantic analysis ability in large-scale environment. Image annotation is an important topic in the semantic analysis of image content. Annotation can establish a classification relationship between image content and semantics. In order to solve the problem of extracting a large amount of data in large-scale image analysis, a content semantic image content analysis and storage scheme based on intelligent computer learning image annotation is proposed. Combined with DSTH work, the program introduces deep learning, visual lexicon and map metadata. Hash semantic metadata supplemental metadata is obtained through deep learning, and semantic metadata is constructed and managed in a hierarchical structure. In addition, according to the characteristics of the graph structure, by improving the PageRank algorithm, the SemRank node ranking algorithm based on Hamming distance is proposed. Experimental results demonstrate the effectiveness and reliability of the algorithm. [ABSTRACT FROM AUTHOR]

The precision of price appraisal is directly linked to the sound development of a market-oriented economy and public benefit. Market comparison approach is a widely used price-appraisal algorithm. However, when the sample size is small, the computing precision of the market comparison approach is severely impaired. To realize precise, intelligent computing under the conditions of a small sample size and a low degree of proximity between samples and the object to be evaluated, this study adopted grey correlation and fuzzy mathematics to optimize the market comparison approach for price appraisal. The proposed algorithm used grey correlation analysis to quantify the weight of various attributes influencing price appraisal. The price correlation between samples and the object to be evaluated was confirmed. The fuzzy mathematical method was then used to obtain the relative weight of samples. The proposed intelligent price-appraisal algorithm was constructed to improve the market comparison approach. Finally, several residences in Hangzhou, Zhejiang province, China were chosen for an empirical analysis and for the verification of the feasibility of the improved price-appraisal algorithm. Results suggest that (1) The mean error of the algorithm is only 0.99%, indicating high computing precision; (2) When the number of objects to be evaluated is large, a computer system can be used to aid the intelligent computing and to efficiently perform intelligent computing. In conclusion, the proposed algorithm can maintain high computing precision under the conditions of a small sample size and a low degree of proximity, and the algorithm can be used as a theoretical basis to realize intelligent mass appraisal. Overall, the proposed algorithm is worthy of further development because it is highly operational. [ABSTRACT FROM AUTHOR]

This work aims to improve the accuracy of financial crisis early warnings for small and medium-sized enterprises (SMEs) to help them uncover hidden dangers during the latent period of a crisis and to respond effectively. First, this work analyses the development characteristics of SMEs, the crises they face, and a financial crisis early warning system. Then, a statistical method is used to determine which indicators are significant to implement the model for an early warning financial crisis to determine the index system. To do this, relevant machine learning (ML) algorithms are introduced to realize a corporate financial crisis management and early warning system (CFCM-EWS) using the stacking fusion method. The CFCM-EWS can mine relationships among the data and analyze nonlinear and difficult-to-explain problems. A comparison of classical financial early warning models (backpropagation neural network (BPNN) and logistic regression models) with the XGBoost model highlights the advantages of the XGBoost model. The model's discrimination results show that its prediction accuracy for special treatment and nonspecial treatment corporates is 85.8 % and 81.9 %, respectively. The logistic regression, XGBoost, and BPNN models are fused using the stacking method. ML provides a more practical prediction method with higher efficiency and accuracy than does traditional econometric models. The final fusion model outperforms the voting and averaging methods in prediction performance. The CFCM-EWS based on intelligent computing discussed here is of great value for SMEs to accurately predict financial crises and adopt timely countermeasures. • Established a financial crisis early warning system for the crisis of SMEs. • For CFCM and early warning systems, this study used the Sacking fusion method. • The voting and averaging methods are used for prediction performance. • CFCM early warning System is significant for taking countermeasures promptly. [ABSTRACT FROM AUTHOR]

In this investigation, a comparative experimental study is conducted to obtain an efficient hybrid intelligent framework for analysing and optimising the operating parameters of the laser solid freeform fabrication (LSFF) process. Here, the experimental studies are conducted in two different stages. In the first stage, different concepts of machine learning systems are taken into account to find a simple yet accurate intelligent model for identifying the LSFF process. To do so, multi-layered neural network with different types of analytical, gradient-based and heuristic learning strategies, i.e. extreme learning machine, back-propagation and steepest descend gradient-based learning and Nelder–Mead simplex heuristic, respectively, are adopted and applied to the LSFF process. In the second stage, different types of swarm- and evolutionary-based metaheuristics, i.e. differential evolutionary algorithm, particle swarm optimisation, the great salmon run, firefly algorithm, bee algorithm, are used to simultaneously find the optimal values of melt-pool depth and clad height during the LSFF process. The statistical results of the simulation indicate that the conducted experiments can result in a fast, robust and accurate hybrid intelligent system which can easily cope with the nonlinearities and uncertainties of the resulting optimisation problem. [ABSTRACT FROM AUTHOR]

Human activity recognition and classification are some of the most interesting research fields, especially due to the rising popularity of wearable devices, such as mobile phones and smartwatches, which are present in our daily lives. Determining head motion and activities through wearable devices has applications in different domains, such as medicine, entertainment, health monitoring, and sports training. In addition, understanding head motion is important for modern-day topics, such as metaverse systems, virtual reality, and touchless systems. The wearability and usability of head motion systems are more technologically advanced than those which use information from a sensor connected to other parts of the human body. The current paper presents an overview of the technical literature from the last decade on state-of-the-art head motion monitoring systems based on inertial sensors. This study provides an overview of the existing solutions used to monitor head motion using inertial sensors. The focus of this study was on determining the acquisition methods, prototype structures, preprocessing steps, computational methods, and techniques used to validate these systems. From a preliminary inspection of the technical literature, we observed that this was the first work which looks specifically at head motion systems based on inertial sensors and their techniques. The research was conducted using four internet databases—IEEE Xplore, Elsevier, MDPI, and Springer. According to this survey, most of the studies focused on analyzing general human activity, and less on a specific activity. In addition, this paper provides a thorough overview of the last decade of approaches and machine learning algorithms used to monitor head motion using inertial sensors. For each method, concept, and final solution, this study provides a comprehensive number of references which help prove the advantages and disadvantages of the inertial sensors used to read head motion. The results of this study help to contextualize emerging inertial sensor technology in relation to broader goals to help people suffering from partial or total paralysis of the body. [ABSTRACT FROM AUTHOR]