Paper Title

THE IMPACT OF SOIL POLLUTION ON HUMAN HEALTH AND THE ENVIRONMENT

Abstract

The study of soil pollution has occupied the attention of a large number of researchers because of its continuity and effect on humans, animals and plants alike. Soil pollution occurs as a result of the entry of elements that change the composition and organism of the soil, and reduce its fertility, making it more vulnerable to drought, and unsuitable for agriculture. This study proposes an index system for evaluating the inherent risk level of polluting factories and introduces an integrated risk assessment method based on human health risk. This study will discuss the scientific and technical issues analysts should consider regarding the use of these methods to estimate the burden of disease attributable to outdoor fine particulate pollution in their own settings. Most of the agricultural lands are irrigated by rivers polluted with factory wastewater. Also, the research discusses the most important sources of pollution represented by pesticides and chemical fertilizers that contain toxic substances and seep into the soil to remain for a long time, and contribute to eliminate soil fertility. The research deals with the impact of logging the desertification of agricultural areas and their transformation into a vast desert that is unsuitable for agriculture, and pollutes the soil with hydrocarbons. Attention was drawn to the problem of existing inequalities in the global burden between economies with different income levels.
KEYWORDS : soil pollution, environmental factors, rivers polluted, factory wastewater, soil fertility, wastewater, human health risk.

Paper Title

A STUDY ON DEVELOPMENT OF A BIOMETRIC AUTHENTICATION SYSTEM

Abstract

This study covers the field of biometric systems focus on biometric authentication systems. A short and general overview of biometric authentication systems gives some insight in how the various biometric data can be used for authentication. The problems of varying biometric data, caused by noise respectively human nature and approaches to solve these problems with multi-biometric systems in combination with information fusion, are also discussed. Since there is such a vast range of variances for the usage of biometric systems, some type of statistics is determined. Biometrics is the measurement of biological data. The term biometrics is commonly used today to refer to the authentication of a person by analyzing physical characteristics, such as fingerprints, or behavioral characteristics, such as signatures. Since many physical and behavioral characteristics are unique to an individual, biometrics provides a more reliable system of authentication than ID cards, keys, passwords, or other traditional systems. The word biometrics comes from two Greek words and means life measure. To provide a comprehensive survey, we not only categorize existing biometric techniques but also present detailed of representative methods within each category. Physical characteristics commonly used in biometric authentication include face, fingerprints, handprints, eyes, and voice. Another program compares this pattern with the one representing the user that was recorded earlier and stored in the system database.
KEYWORDS : Biometric authentication, biometric systems, fingerprints, handprints, eyes, and voice, computer networks, database, ID cards, keys, passwords.

Paper Title

A STUDY ON IRIS RECOGNITION CHALLENGES AND MACHINE LEARNING ALGORITHMS FOR UNIMODAL BIOMETRIC

Abstract

Iris Recognition is one of the most challenging and fastest growing areas in the field of biometrics. This study focuses about the brief summary of iris recognition system for identification and verification of iris images. The fusion of biometrics leads to security systems that exhibit higher recognition rates and lower false alarms compared to unimodal biometric security systems. Supervised learning was carried out using a number of patterns from a well-known benchmark biometrics database, and the validation/testing took place with patterns from the same database which was not included in the training dataset. Furthermore, this research might explore the adaptability of these systems to real-world scenarios, addressing challenges such as variations in lighting conditions, occlusions, and robustness against spoof attacks. Subsequently, machine learning algorithms such as Convolution Neural Networks (CNNs), Support Vector Machines (SVMs), or Deep Learning models are applied to classify and identify individual irises. We comprehensively evaluated our algorithm on three publicly available iris databases for which the results proved satisfactory for real-time iris recognition. The fusion of biometrics leads to security systems that exhibit higher recognition rates and lower false alarms compared to unimodal biometric security systems.
KEYWORDS : Machine learning algorithms, benchmark biometrics database, Support Vector Machines (SVMs), iris databases, Convolution Neural Networks (CNNs).

Paper Title

Kum Shanti Devi Bais

Abstract

The degree of non-ideality in non-ideal liquid-gas mixtures must be determined. Several quantitative qualities may be utilized to quantify surplus properties experimentally. These may be approximated by various model equations. Density, surplus molar volume, and apparent molar volume define a binary mixture. This and other volumetric properties may be found in thermodynamics. Volumetric and thermodynamic properties depend on temperature and pressure. Partial molar volumes, partial excess molar volumes, partial molar volumes at infinite dilution, isothermal coefficients of pressure, partial molar enthalpy, thermal expansion coefficients, and excess thermal expansion coefficients all show a mixture\'s thermodynamic significance. This review covers mixture thermodynamics research.
KEYWORDS : Infinite dilution, excess properties, enthalpy, partial properties

Paper Title

LOAN APPROVAL PREDICTION BASED ON MACHINE LEARNING

Abstract

Modern financial landscape, efficient and accurate loan processes are critical for managing risk and enhancing customer satisfaction. This study explores the application in machine learning techniques to predict loan approval outcomes. Analyzing historical loan data, including applicant demographics, financial history, and other relevant features, various machine learning algorithms are logistic regression, decision trees, random forests, and gradient boosting are employed to develop predictive models. Performance of these models is evaluated using metrics like accuracy, precision, recall, and with a focus on identifying the most effective approach for predicting loan approvals. Machine learning models can significantly improve the precision of loan approval predictions, leading to more informed decision-making and optimized resource allocation. This research highlights the potential for machine learning to enhance financial decision processes and proposes future directions for integrating these techniques into practical loan approval systems.
KEYWORDS : Machine Learning, logistic regression, decision trees, random forests.

Paper Title

IMPLEMENTING EFFICIENT MACHINE LEARNING TECHNIQUES TO IDENTIFY COCONUT LEAVES ENHANCES THE DIAGNOSTIC ACCURACY OF DISEASES

Abstract

Reducing substantial losses in coconut plantations requires precise and prompt disease prediction for coconut palms. Diseases such as deadly yellowing, bud rot, root wilt, and stem bleeding pose significant threats to coconut farming, a vital agricultural pursuit in tropical areas. It is crucial to detect and manage these diseases early on because they cause devastating crop destruction and economic loss. Climate change, ineffective resource management, and a lack of disease preventive measures are among environmental variables that endanger the long-term viability of coconut cultivation. One potential resolution is the application of ML and predictive modelling to precision agriculture. In order to forecast when diseases may manifest and propose treatments in advance, these models can combine data from a wide variety of sources, such as image-based analysis, environmental parameters, and feedback from sensors. More proactive steps may be taken by farmers with the help of this technology, which improves agricultural productivity and sustainability in the long run
KEYWORDS : Machine Learning, Deep Learning and coconut Palm, disease prediction and RNN

Paper Title

DIABETES-RELATED RETINAL DEGENERATION PREDICTION USING A CLASSIFICATION TECHNIQUE

Abstract

Diabetes is a chronic condition characterised by elevated blood glucose levels due to insufficient insulin production or the body\'s ineffective use of insulin. Effective management of diabetes is crucial for preventing complications such as diabetic retinopathy, foot ulcers, and cardiovascular issues. Image processing and deep learning exhibit considerable promise in improving diabetes treatment through early detection, monitoring, and therapy optimisation. The administration of diabetes by image processing and deep learning provides significant advancements in the early identification and surveillance of diabetic complications, including diabetic retinopathy and foot ulcers. Current investigations into non-invasive glucose monitoring and tailored medication optimisation suggest that deep learning methodologies have the potential to revolutionise diabetes management, improve patient outcomes, and reduce healthcare costs. However, challenges in data accessibility, model interpretability, and clinical integration must be addressed to fully realise these benefits.
KEYWORDS : Image processing, deep learning, Diabetic Retinopathy and CNN