http://repository.bothouniversity.ac.bw:8080/buir/handle/123456789/23 2026-04-13T11:45:46Z 2026-04-13T11:45:46Z Mbunge, Eillot Batani, John Moyo, Enos http://repository.bothouniversity.ac.bw:8080/buir/handle/123456789/273 2025-04-07T15:37:05Z 2024-01-01T00:00:00Z

Towards Data-Driven Artificial Intelligence Models for Monitoring, Modelling and Predicting Illicit Substance Use Mbunge, Eillot; Batani, John; Moyo, Enos . Illicit substance use (ISU) is a major public health problem and a significant cause of morbidity and mortality globally. Early assessment of risk behaviour, predicting, identifying risk factors, and detecting illicit substance use become imperative to reduce the burden. Unfortunately, current digital tools for early detection and modelling ISU are largely ineffective and sometimes inacces sible. Data-driven artificial intelligence (AI) models can assist in alleviating the burdenandtacklingillicit substance use buttheir adoption anduseremainnascent. ThisstudyappliedthePRISMAmodeltoconductasystematicliteraturereviewon the application of artificial intelligence models to tackle illicit substance use. The study revealed that elastic net, artificial neural networks support vector machines, random forest, logistic regression, KNN, decision trees and deep learning mod els have been used to predict illicit substance use. These models were applied to tackle different substance classes, including alcohol, cannabis, hallucinogens, tobacco, opioids, sedatives, and hypnotics among others. The models were trained and tested using various substance use data from social media platforms and risk factors such as socioeconomic and demographic data, behavioural, phenotypic characteristics, and psychopathology data. Understanding the impact of these risk factors can assist policymakers and health workers in effective screening, assessing risk behaviours and, most importantly, predicting illicit substance use. UsingAImodelsandriskfactorstodevelopdata-drivenintelligentapplicationsfor monitoring, modelling, and predicting illicit substance use can expedite the early implementation of interventions to reduce the associated adverse consequences.

2024-01-01T00:00:00Z G. Bharathi Mohan1 ·R. PrasannaKumar1 ·P. VishalKrishh1 ·A. Keerthinathan1 · G. Lavanya1 ·Meka Kavya Uma Meghana1·Sheba Sulthana1·Srinath Doss2 http://repository.bothouniversity.ac.bw:8080/buir/handle/123456789/270 2025-04-07T12:28:44Z 2024-01-01T00:00:00Z

Ananalysis of large language models: their impact andpotential applications G. Bharathi Mohan1 ·R. PrasannaKumar1 ·P. VishalKrishh1 ·A. Keerthinathan1 · G. Lavanya1 ·Meka Kavya Uma Meghana1·Sheba Sulthana1·Srinath Doss2 Large language models (LLMs) have transformed the interpretation and creation of human language in the rapidly developing field of computerized language processing. These mod els, which are based on deep learning techniques like transformer architectures, have been painstakingly trained on massive text datasets. This study paper takes an in-depth look into LLMs,includingtheirarchitecture,historicalevolution,andapplicationsineducation,health care, and finance sector. LLMs provide logical replies by interpreting complicated verbal patterns, making them beneficial in a variety of real-world scenarios. Their development and implementation, however, raise ethical concerns and have societal ramifications. Under standing the importance and limitations of LLMs is critical for guiding future research and ensuringtheethicaluseoftheirenormouspotential.

2024-01-01T00:00:00Z Yamuna Devi, M. M., J. Jeyabharathi, S. Kirubakaran, Sreekumar Narayanan, T. Srikanth, and Prasun Chakrabarti. http://repository.bothouniversity.ac.bw:8080/buir/handle/123456789/261 2024-05-17T13:19:44Z 2023-01-01T00:00:00Z

Efficient segmentation and classification of the lung carcinoma via deep learning. Yamuna Devi, M. M., J. Jeyabharathi, S. Kirubakaran, Sreekumar Narayanan, T. Srikanth, and Prasun Chakrabarti. Lung malignancy represents a group of diseases that affect people worldwide. According to the reports, 1.69 million people died in 2015. Presymptomatic work raises patient ability and boosts treatment success. The accuracy of disease detection, velocities, and computer technology standards are used to calculate CAD systems. This study investigated the malignant tumor detection method over a few currently used structures. This article discusses lung carcinoma segmentation and the classification of methods. The five steps of the computer-assisted workflow are image acquisition, preprocessing, segmentation, feature extraction, and classification. It focuses on the edge-segmentation process, which is becoming more popular for effective image segmentation in regions. Moreover, the key attributes can be deduced from feature resemblance and transmitted to the classification technique. The lung cancer image's area features are classified using a clustering technique. The lung cancer image's cancer field has been removed using CNN. Moreover, the histogram and adaptive median filter are applied to enhance segmentation performance. The experimental studies utilized basic images acquired from the database and existing health data obtained from the patient. The results demonstrated that the proposed statistical method's performance, which can produce better results than other existing predictions, is superior.

2023-01-01T00:00:00Z Viswanathan, Sankaranarayanan Vengatesh, Kumar http://repository.bothouniversity.ac.bw:8080/buir/handle/123456789/255 2023-07-11T12:18:15Z 2023-02-01T00:00:00Z

A Novel Approach for Analysis and Prediction of Students Academic Performance Using Machine Learning Algorithms Viswanathan, Sankaranarayanan; Vengatesh, Kumar Educational data mining has become an efective tool for exploring the hidden relationships in educational data and predicting students’ academic performance. The prediction of student academic performance has drawn considerable attention in education. However, although the learning outcomes are believed to improve learning and teaching, prognosticating the attainment of student outcomes remains underexplored. To achieve qualitative education standard, several attempts have been made to predict the performance of the student, but the prediction accuracy is not acceptable. The main purpose of this research is significantly predict the student performance to improve the academic results. In order to accomplish the prediction with supplementary exactness, XGBoost based methods have been adopted. This work introduces a novel hybrid Lion-Wolf optimization algorithm to solve the problem of feature selection. Two level overlap improves the exploitation part. First phase overlap is used for feature selection and second phase used for adding some more important information and improve the classification accuracy. The XGBoost classifier improved the classification accuracy which is most famous classifier based on wrapper method. XGboost model using two different parameter adjustment methods are compared. XGBoost based on hybrid Lion-Wolf optimization performs better than traditional XGBoost on training accuracy and efficiency. Experiments are applied using the dataset and results prove that proposed algorithm outperform and provide better results

2023-02-01T00:00:00Z