This study is an early-bird attempt to probe the reasons behind Cash Holding among Indian urban local bodies (ULBs). The risk of uncertainty, volatility in cash flows, and dynamic economic conditions insist on the management of ULBs viewing cash as a strategic tool for situation management and scenario planning. This paper explores the reasons for Cash Holdings by analyzing 38 Indian municipal corporation’s data over 5 years. The study results revealed the Existence of a high degree of inverse relationship between municipal Cash Holdings and aspects such as Growth, Size, State Revenue, Capital Expenditure, and Debt Per Capita. In contrast, Administrative Expenses and Tax Revenue positively correlated with municipal Cash Holdings. The study results empirically identified the variables, namely Size, Administrative Expense, State Revenue, and Debt Per Capita, as the primary reasons for holding cash by Indian municipal corporations and pointed out the Existence of differences among different categories of municipal corporations in India.
The Marine Predators Algorithm, known for its simplicity and efficiency, is a widely used optimization method. However, it is prone to structural bias, which can cause the algorithm to favor specific areas in the search space without considering the objective function. This bias can hinder exploration, leading to the population repeatedly visiting certain locations without gaining new information, resulting in increased computational burden. In this study, we extensively investigated the occurrence and types of structural bias in the Marine Predators Algorithm. We also assessed newly developed algorithm variants, such as the Opposition-based Local Escaping Marine Predator Algorithm, to determine their susceptibility to structural bias. To detect and analyze structural bias and its types, we employed a straightforward yet effective methodology called the signature test. Through a comprehensive analysis, we have identified algorithms that exhibit unbiased behavior. We believe that our analysis will serve as a valuable resource for practitioners interested in analyzing the theoretical aspects of their algorithms.
Multi-modal medical image fusion plays a crucial role in improving clinical decision-making and boosting diagnostic accuracy by combining comprehensive features from more than one modality. This paper explores the medical image fusion of three different modalities—MRI-gad/MRI-T1/MRI-T2, PET, and CT using three different low-frequency fusion rules—random weighted average, sliding window, and maximum selection. Each rule is evaluated based on its ability to preserve the most salient features from the source images. This study also presents an optimized weighted average fusion rule for the high-frequency coefficients. This approach employs the three most potent metaheuristic optimization algorithms—Differential Evolution, Particle Swarm Optimization, and Genetic Algorithm. These optimization methods aim to dynamically adapt the fusion weights to guarantee the most pertinent and instructive features in the final fused image. The outcome of this study is to present a comparative analysis for the advancement of multi-modal medical image fusion, contributing to improved medical diagnostics, treatment planning, and disease management.
This research paper examines the impact of customer segmentation on purchasing intentions in online shoppers. Using the K-means method and the Online Shoppers Purchasing Intention Dataset, which includes 12,330 sessions from a popular e-commerce website, the study analyzes customer behavior. The findings highlight notable sales surges in August, varying preferences across countries, and the significance of customer engagement and transactional activity. Through RFM and K-means segmentations, exceptional customers with high engagement and value are identified, emphasizing the importance of customer retention strategies. Moreover, opportunities to re-engage less active customers are uncovered. The research provides valuable insights for businesses to optimize marketing efforts, enhance customer satisfaction and loyalty, and maximize revenue potential in the online shopping.
The Grasshopper Optimization Algorithm (GOA) is a relatively recent population-based stochastic approach for solving the nonlinear global optimization problems. A number of attempts have been made in the past to improve the efficiency of population-based approaches by combining them with the features of other approaches. In this paper, an attempt has been made to introduce an enhanced version of GOA by combining it with another population-based approach that is the Self-Organizing Migrating Algorithm (SOMA). GOA is combined with SOMA, and a hybrid variant, SOMGOA, is proposed. The effectiveness of this approach is analyzed on the basis of results, and comparative analysis is made against the previously published results by other algorithms on 15 standard benchmark functions. It is concluded that SOMGOA outperforms all and can be used further to solve nonlinear optimization problems.
An increase in global population growth has necessitated an increase in food production. One of the main factors influencing annual agricultural production is the abnormal physiological functioning of plants or plant diseases, which directly affects the vegetation, leading to a reduction in plant yields and in the worst case, may even destroy the entire plantation. A majority of the diseases can be identified through the plant’s leaves, which a plant pathologist traditionally does. This, however, is a time-consuming task and the accuracy of diagnosis depends a lot on the expertise of the pathologist. Convolution neural networks (CNNs) have shown immense potential in image identification tasks. However, optimizing its hyperparameters and layouts is a challenging task. We proposed a genetic algorithm to enhance the performance of CNNs for plant disease identification by assessing the most effective hyperparameters and architecture for the fully connected layers of four cutting-edge CNNs: VGG16, Xception, DenseNet201, and ResNet152V2. The results show that genetic algorithms possess the potential to enhance the performance of CNN architecture in the agriculture domain, especially when used for plant disease identification.
The integration of Artificial Intelligence (AI) and Machine Learning (ML) is profoundly impacting the civil and criminal legal systems, particularly in India, where over 5 crore cases are pending across courts, including 59 lakh in High Courts, as per the National Judicial Data Grid (NJDG). While the judiciary has historically lagged in digital adoption, accelerated efforts, especially spurred by the COVID-19 pandemic’s push for e-filing and online hearings, are driving a gradual but significant shift towards e-governance. This digital transformation, supported by initiatives like India’s e-Courts Project, aims to enhance efficiency, accessibility, and transparency. Globally, AI/ML is being adopted for diverse applications, including smart case management, improved investigation through systems like CCTNS, predictive analytics for insights into case outcomes, streamlined legal research with tools for finding precedents and summarizing judgments (like India's SUVAS), and supporting decision-making in areas like evidence analysis and sentencing guidelines. However, this promising future necessitates careful consideration of ethical challenges such as algorithmic bias from flawed training data, the black box problem of opaque AI decision-making, accountability for AI-induced errors, and robust data privacy and security measures. The ultimate goal is to leverage AI and digitization to democratize justice delivery, making it more affordable and accessible for the public while enabling the State to deliver justice efficiently and transparently.
In the present paper, a case study of selected Indian Pulp and Paper mills is discussed where the aim is to measure the efficiency of the mills based on the available data containing information about the 32 pulp and paper mills as per the 6 selected criteria. To calculate the efficiency of each Mill, data envelopment analysis (DEA) is an appropriate technique. However, classical DEA models often neglect the presence of undesirable outputs or inputs that need to be minimized or reduced. This study addresses this limitation by developing a two-stage network DEA framework that contains undesirable variables. A two-stage network DEA is developed for evaluating the efficiency of decision-making units (DMUs) operating in a multi-stage process. Here, a DMU represents a Pulp and Paper mill. The results highlight the importance of considering undesirable variables in assessing the efficiency of DMUs accurately. The developed model offers valuable insights for decision-makers by identifying areas for improvement and suggesting strategies to enhance efficiency while dealing with undesirable variables. The efficiency scores obtained through the two-stage network DEA technique are compared with those calculated by the classical DEA method.
Topic modeling offers a useful way to examine the topic labels of extensive document collections, facilitating the organization and outline of the themes within that collection. Previous researchers have suggested considering the probabilistic model, where each document is the convex combination of topic vectors, and the topic vector is a distribution of words. However, finding an appropriate distribution vector for each topic is not easy for a high-dimensional word co-occurrence space. This work provides an alternative topic vector inference method combined with non-negative matrix factorization for learning high-quality topics. To verify the effectiveness and priority of the proposed method, we experiment with three public benchmark datasets, NIPS, Movies, and NYtimes, and show a competitive performance.
The use of recycled waste is increasing nowadays, and it is a major concern to use recycled post-consumer waste in food packaging. The use of recycled waste will reduce waste and will contribute to sustainability and a circular economy. In this study, a comparison has been made between recycled and virgin Polypropylene (PP) materials that have been used and going to be used in food packaging respectively. Herein, virgin and recycled PP samples were analyzed, and hundreds of Volatile Organic Compounds (VOCs), odorous, and semi-VOCs have been observed with the help of Gas chromatography-mass spectrometry (GC-MS). These samples were analyzed two times to get high efficiency. To classify the VOCs and odorous compounds within the Virgin Polyethylene (Vpet) and Recycled Polyethylene (Rpet) classes, four machine learning algorithms were applied: Random Forest (RF), XGBOOST, Support Vector Machine (SVM), and Gradient Boosted Decision Tree (GBDT). Among these algorithms, Random Forest achieved the highest accuracy. Additionally, the Mean Decrease Impurity method was utilized to determine the feature importance in the classification process.
Acoustic levitation uses sound waves to counteract gravity, create standing waves, and then keep objects in air. In this paper, we will start with an introduction about acoustic levitation, its various types, introduction about different parts, their work in this project, and step-by-step process to make a working model of acoustic levitation. Acoustic levitation was discovered a century ago. At that time, it was limited to levitating small objects, but now it can be useful to levitate objects larger than the acoustic wavelength. It is helpful to trap an assortment of materials like fluids, solids, and living things like, holding insects or other very delicate living things allows us to effectively study them under a magnifying lens without contacting them. It’s useful in many other places like the mobile industry, packaging industry, etc.
The crime rate is increasing daily, and finding criminals in such a large population will be challenging. It is known that the face is a unique and determining part of the human body that identifies a person; hence, it can be used to track down the identity of a criminal. A solution to this problem would be to create a system that controls CCTV cameras and monitors them 24/7 to identify criminals and notify the nearest police station. In times like now, security cameras can be found almost everywhere, and criminal face recognition systems can be implemented using the previously captured faces from police stations and criminal images. This article proposes a system that can enhance criminal distinction and provide a more effective and efficient outlook for the police department. This proposed system consists of a database where the appearance of the criminal will be uploaded along with the criminal description he has made, and then the database will provide the information to the system. Once the image is in the database, the system will detect the criminal by comparing the captured appearance, which can be done with facial recognition software. The crime data in our database along with the people who come to this public place, if the face of a person from the public place matches the present data in our database, will be the most recent to notify the police department. This leads to improved social security.
The Salp Swarm Algorithm is a popular optimization method known for its simplicity and efficiency. However, it is susceptible to structural bias, which can cause the algorithm to favor specific regions of the search space without regard for the objective function. Structural bias can hamper exploration, leading to the population revisiting certain locations without acquiring new information, which adds to the computational load. This study involves a comprehensive investigation of the occurrence and types of structural bias in the Salp Swarm Algorithm. Additionally, we evaluate two newly developed variants of Salp Swarm Algorithm, namely the Laplacian Salp Swarm Algorithm and the Quadratic Approximation Salp Swarm Algorithm, for their structural bias. To detect and analyze structural bias and its type, a simple yet effective methodology called the signature test is employed. After conducting a thorough analysis, we have identified algorithms that have demonstrated unbiased behavior. We anticipate that our analysis will be a valuable resource for practitioners who are interested in analyzing the theoretical aspects of their algorithms.
Society 5.0 represents a transformative era where the convergence of technology and society reshapes decision-making processes. This paper explores the challenges and opportunities in decision-making within Society 5.0 and focuses on the Sustainable Development Goals (SDGs) as a critical framework for sustainable development. Assessing the progress of Indian states toward these goals is crucial for effective policy formulation. To comprehensively analyze the performance of Indian states in achieving the SDGs, this study employs the TOPSIS method and cluster analysis. Integrating these approaches establishes a robust framework for benchmarking and evaluating state performance. The findings provide valuable insights into progress variations, identifying improvement areas and enabling targeted policy interventions and resource allocation. This research enhances our understanding of sustainable development progress at India’s state level. By informing policy decisions and fostering effective strategies, it contributes to the successful implementation of the SDGs. The manuscript presents the methodology, data analysis, and results, providing a structured assessment of Indian states’ SDG performance. Through this study, we aim to support evidence-based decision-making and promote sustainable development in India.
Inventory management is the process of maintaining an account of company’s products that are being ordered, sold, and stored. This process comprises management of raw materials, processing, and warehousing of finished products. The practice analyzes and retorts to tendencies to ensure there’s always sufficient stock to satisfy customer orders and appropriate warning for shortage. A well-managed inventory helps in saving money, improving cash flows, and satisfying customers. An automated system for this limits the danger of blunder. The implementation of such system would reduce the work done by humans to about 90
The ability to recognize emotions in speech has the potential to enhance a variety of areas, including safety, customer service, mental health, and communication. Speech-based emotion detection focuses on classifying audio recordings according to specific emotions. The frequency and pitch are read from the audio files. We are contrasting the accuracy of the two models, the Multi-layer Perceptron Classifier model and the Support Vector Classification model, in order to attain the goal of recognizing the fundamental emotions, such as calm, happy, sad, and angry. We used the Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS) dataset with MEL-Frequency Cepstral Coefficients, Chroma, and MEL as feature extraction techniques. The outcomes demonstrate that the MLP classifier outperformed the SVC model. For the MLP and SVC models, 92.34
Sugarcane is a major commercial crop grown in many countries around the world. It is an important source of income for millions of farmers and workers involved in the sugar industry. The cultivation and processing of sugarcane also create jobs in other industries such as transportation, manufacturing, and retail. Additionally, sugarcane is used not only for sugar production but also for biofuel production, making it a versatile and valuable crop. In this article, we will explore the current state of sugarcane production worldwide, including the top-producing countries. The country-wise ranking of sugarcane is determined with respect to five criteria using Multi-Criteria Decision-Making (MCDM) methodology. The final ranking is compared through three MCDM methods viz. VIKOR, TOPSIS, WSM.
Big Data and Cloud Computing have simplified many business operations together along with this these technologies are worrying the IT industry also. Every day several sources generate huge amounts of data. The data is so big that it cannot be processed using conventional methods. Statistics show that more than 44 zettabytes of data are generated every day. Every human generates 1.7 MB of data per second. With the help of cloud computing, these processes can be carried out efficiently. Many sectors like banking and financial services are using big data analytics. According to the latest approximations, 328.77 million terabytes of data are created each day. This paper endeavors to analyze big data which exhibits numerous benefits in multiple domains and industries such as education, wellness, and businesses and concluding with a case study. It has also witnessed a change in how data is handled and examined as a result of interaction in the middle of substantial Data and Cloud Computing.
The Bidirectional Encoder Representations from Transformers (BERT) algorithm has become a valuable tool for learning common languages and has generated innovative outcomes on numerous projects. Social media stages offer a resource of user-generated information that can be analyzed to pick up bits of knowledge about human behavior and assumption. In this survey paper, we offer an outline of the application of the BERT calculation to a social media investigation. This paper is about the BERT algorithm and how its work is also included in preprocessing and tokenizing social media information. We also discuss the methods used to fine-tune BERT models for social media-specific datasets. We also investigated the diverse estimation approach and named substance acknowledgment utilizing BERT. At last, this paper presents the confinements of the BERT algorithm for social media investigation and long-standing time bearings for investigation in this field. By and large, this audit highlights the potential of the BERT calculation as an important device for social media examination and its capacity to supply unused experiences into human behavior and estimation.
