Volume 12 | Issue 4 |

 Abstract

The implementation of a back-propagation neural network to classify the MNIST handwritten digit database is shown in this research. Here, we assess the neural network's performance using the plot of loss during training and classification accuracy. The outcomes of the experiment indicate that back-propagation neural networks can be used to real-world categorization problems to a certain extent. Additionally, we attempt to use Autoencoder to produce picture compression and analysed its results. Reducing the size of the network helps to speed up the neural network and enhance performance. Though several causes are listed, the accuracy rate cannot be guaranteed. Next, we attempt to convert the original neural network into a Convolutional Neural Network (CNN) by changing its structure. According to the findings, CNN can be utilized to enhance performance while tackling image identification tasks. Additionally, we build a Conv Autoencoder structure by combining CNN and Autoencoder, and we run various tests.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

back-propagation neural network, MNIST handwritten digit database, neural network performance, loss plot, classification accuracy, Autoencoder, picture compression, network size reduction, Convolutional Neural Network (CNN), image identification tasks, Conv Autoencoder structure, performance enhancement, real-world categorization, neural network experiments, neural network analysis.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

For long ago, biometric authentication has been performed for high-security applications like bank lockers, private places. Here studies made on the physiological attributes of an individual can be represented with his finger print, iris, etc. Among those finger vein identification is a new methodology in the biometric recognition system. Here, finger vein patterns can be used to give authentication to an individual for accessing high- security applications. In this paper, we are going to perform finger vein recognition Quadrature Discriminant analysis method. Finger vein images are subjected to preprocessing steps for making the image more stable for further processing. Then QDA process has been applied, followed by the Minimum Distance Classifier.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Biometric recognition, finger vein, QDA, MDC,

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The surge in online shopping has led to a growing environmental concern: the impact of returned items. This paper explores the concept of sustainable retail returns, aiming to balance customer convenience with environmental responsibility. The growing popularity of online shopping has led to a surge in retail returns, creating significant environmental and economic concerns. This paper explores the concept of sustainable retail returns, analyzing strategies that minimize the environmental impact of the returns process while maintaining customer satisfaction and profitability.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Traditional returns, significant waste, emissions, resource depletion, sustainable return strategies, customer satisfaction and profitability

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

In response to the dynamic challenges posed by contemporary cloud environments, the development of a Cloud Security Posture Management (CSPM) tool gains significance. This paper outlines the systematic methodology employed in creating a CSPM tool without reliance on machine learning. The process begins with a meticulous analysis of requirements, followed by the design of a scalable architecture ensuring seamless integration with major cloud platforms. Automated assessments are crafted to evaluate security configurations against established best practices and compliance standards, offering real-time monitoring and remediation. The user interface is thoughtfully designed for ease of use by security professionals. Continuous improvement mechanisms, including regular updates and adaptability to evolving cloud technologies, are prioritized. Through comprehensive testing and deployment, the CSPM tool emerges as a robust solution to proactively address security challenges without the inclusion of machine learning components.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Amazon Web Services (AWS), Azure, Cloud Security Posture Management (CSPM), Cloud Security

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Some of the Ayurvedic classics and texts in medieval period have classified all the poisons in to two categories as Mahavisha and Upavisha based on their toxicity and potency. Upavisha are the group of drugs which were less toxic in nature but produce certain toxic symptoms on consumption or administration. Nerium oleander most commonly known as oleander or nerium, is a shrub or small tree cultivated worldwide in temperate and subtropical areas as an ornamental and landscaping plant. It is the only species currently classified in the genus Nerium, belonging to subfamily Apocynoideae of the dogbane family Apocynaceae. Nerium contains several toxic compounds, and it has historically been considered a poisonous plant. However, its bitterness renders it unpalatable to humans and most animals, so poisoning cases are rare and the general risk for human mortality is low. So, this article is focusing on the Nerium upvisha mentioned Agadtantra of Ayuveda.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Nerium indicum, upvisha.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This study examines consumer perception and preference towards Maruti Suzuki's marketing strategies in the Indian automobile industry. Employing both qualitative and quantitative methods, it explores various factors influencing consumer attitudes, such as branding awareness, product quality perception, and the impact of promotional efforts. By considering demographic variables, the research aims to uncover insights to guide Maruti Suzuki's marketing decisions and contribute to the broader understanding of consumer behaviour in the Indian automotive market.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Marketing Strategy, Respondent perception

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The smart garbage management is a mobile application designed to revolutionize waste management practices by leveraging modern technology. With the proliferation of smart phones, there exists an opportunity to optimize waste collection processes and enhance efficiency in urban environments. This application offers a user-friendly interface for both citizens and waste management authorities to streamline the waste collection process. Through smart bin technology, users can able to post the bin status like filled. Admin can monitor the status of the bin by the user status update. Once the user updates the bin status then the admin allots a driver for that particular bin. The driver once completed the work will update the status as completed. The Smart Garbage Management Application aims to promote efficient waste management practices, reduce environmental impact, and enhance the overall quality of life in urban areas.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Garbage Management, Table, System requirements

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Cybersecurity breaches posture a critical danger to organizations and people alike, requiring the advancement of progressed precise and explanatory apparatuses to defend against cyberattacks successfully. In reaction to this squeezing require, this extend presents a comprehensive approach leveraging profound learning techniques to foresee and analyze cyber hacking breaches. The venture unfurls in a few stages, starting with the collection of different datasets including data relating to verifiable cyber hacking episodes. These datasets are subjected to fastidious preprocessing methods to guarantee information keenness and consistency, with particular on dealing with lost values, encoding categorical factors, and normalizing numerical highlights. In this way, highlight building procedures are connected to extricate relevant data that can serve as inputs to profound learning models. Central to the extend is the investigation of different profound learning structures, counting repetitive neural systems (RNNs), convolutional neural systems (CNNs), and transformer-based models, to observe the foremost reasonable demonstrate for the expectation and investigation of cyber hacking breaches. Through iterative experimentation, hyperparameter tuning, and thorough show assessment, the extend endeavors to optimize show execution, endeavoring for accentuation tall precision and unwavering quality in foreseeing cyberattacks. Besides, the venture joins interpretability methods to explain the basic variables driving the model's expectations, in this manner upgrading our understanding of the elements of cyberattacks and encouraging educated decisionmaking. Upon accomplishing palatable demonstrate execution, the extend moves to the sending stage, where the prepared models are coordinates into generation situations for real-time discovery and examination of cybersecurity dangers. This arrangement encourages proactive defense components, empowering organizations to recognize and react quickly to potential breaches, subsequently moderating the affect of cyberattacks. By leveraging cutting-edge profound learning techniques, this extend contributes to reinforcing cybersecurity resistances, bracing advanced frameworks against the revolving scene of cyber dangers.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Cybersecurity, Predictive modeling, Cyber hacking breaches, Malware, Deep learning.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The present study investigates the motivating factors that drive consumers to purchase consumer durables electronic items, commonly referred to as white goods, from online retailers in the regions of Gujarat and Maharashtra. With the burgeoning growth of e-commerce platforms, understanding consumer behaviour and motivations has become crucial for businesses to formulate effective marketing strategies. This research employs a mixed-methods approach, combining qualitative and quantitative techniques to comprehensively analyse the factors influencing consumer purchasing decisions. Quantitative data collection methods include Survey methods, allowing for a nuanced exploration of consumer perceptions, attitudes, and preferences towards online purchasing of white goods. Through thematic analysis, key themes and patterns emerge, shedding light on the underlying motivations guiding consumer behaviour. Complementing qualitative findings, quantitative data is gathered through structured surveys distributed among a diverse sample of consumers in Gujarat and Maharashtra. Statistical analysis techniques, including regression analysis, are utilized to identify significant motivating factors and their impact on consumer buying behaviour. Preliminary findings suggest that convenience, price competitiveness, product assortment, brand reputation, trust in online platforms, and post-purchase services emerge as prominent motivating factors influencing consumers' decisions to buy white goods from online retailers. Furthermore, demographic variables such as age, income level, and prior online shopping experience exhibit varying degrees of influence on consumer preferences and behaviour. The implications of this study extend to both academia and industry. For academia, the research contributes to the existing body of knowledge on consumer behaviour in the context of e-commerce, providing insights into the evolving dynamics of online retailing in the region. For industry practitioners, the findings offer actionable insights for devising targeted marketing strategies, enhancing customer satisfaction, and fostering long-term relationships with online consumers. In conclusion, this study contributes to a deeper understanding of the motivating factors shaping consumer purchasing behavior in the domain of online retailing of consumer durables electronic items/white goods, thereby facilitating informed decision-making for businesses operating in the rapidly evolving e-commerce landscape of Gujarat and Maharashtra.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Consumer behaviour ,Online retailing ,White goods ,Consumer durables ,Motivating factors , E-commerce ,Gujarat, Maharashtra ,Purchase decision ,Consumer preferences

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This study examines student satisfaction with the hiring process in IT companies among placement-registered students in Coimbatore District. Utilizing quantitative methods, structured questionnaires will be distributed to gather data on recruitment procedures, interview experiences, communication effectiveness, and overall satisfaction levels. Findings aim to provide insights for improving recruitment strategies and enhancing student satisfaction in the job placement process.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Placement, IT industry, Coimbatore, College Students

  License

Creative Commons Attribution 4.0 and The Open Definition