The "INDIAai" (National AI Portal of India) portal provides weekly comprehensive articles highlighting the research contributions made by universities and colleges in India. 

We aim to offer thorough reporting on the AI research contributions made by a single institution every week. This series focuses on the quality of research contributions from each Indian institution.

One of the top private universities in the Indian state of Telangana, Woxsen University attained the status of State Private University in 2020. Its component institutions, which include the School of Business, School of Technology, School of Art & Design, School of Law, School of Liberal Arts and Humanities, and School of Architecture and Planning, are housed on a 200-acre residential campus.

Woxsen University's Centre of Excellence is a centre for research and development, emphasizing cutting-edge fields and technology. With an emphasis on supporting industry-driven research and innovation, it offers a forum for knowledge production and cooperation. There are several different CoEs at Woxsen University, including 21 devoted exclusively to students, 26 at the faculty level, and 15 at the university level.

Similarly, Woxsen University's PhD program aims to provide students with the information and abilities to participate in the knowledge-based economy in academic and non-academic fields. Renowned for its PhD program in multiple academic subjects, Woxsen University is an interdisciplinary research-focused university in India. The curriculum places a strong emphasis on demanding coursework. It promotes interaction with peers, mentors, and instructors to develop a thorough understanding of the subject matter and stay current with challenges.

Let us explore the top AI research contributions from Woxsen University, Hyderabad, India.

Performance Analysis of Deep Learning Algorithms in Diagnosis of Malaria Disease

Authors: Hemachandran K, Areej Alasiry, Mehrez Marzougui, Shahid Mohammad Ganie, Anil Audumbar Pise, M. Turki-Hadj Alouane, Channabasava Chola

The research manuscript evaluates deep-learning models for diagnosing malaria from blood smear images. The authors utilized Convolutional Neural Networks (CNN), MobileNetV2, and ResNet50 models on a dataset of 27,558 images from the National Institutes of Health. The study's key finding is that the MobileNetV2 model outperformed others, achieving a 97.06% accuracy rate, which suggests its potential for improving malaria diagnosis in settings with limited healthcare infrastructure. Performance metrics like precision, recall, f1-score, and ROC curve were also analyzed to validate the models' effectiveness.  

Also, it demonstrates the potential of deep learning models, particularly MobileNetV2, to significantly enhance malaria diagnosis through blood smear image analysis. It will facilitate early detection, improving patient outcomes and operational efficiency in the healthcare industry. 

Artificial Intelligence: A Universal Virtual Tool to Augment Tutoring in Higher Education

Authors: K. Hemachandran, Priti Verma, Purvi Pareek, Nidhi Arora, Korupalli V. Rajesh Kumar, Tariq Ahamed Ahanger

The work presents and explores the use of AI to predict the future of higher education. The authors highlight the current challenges in the educational sector, such as faculty and student issues and changing government regulations. They propose a use case model based on student assessment data and employ generative adversarial networks (GAN) and various machine learning algorithms to analyze the data. The study achieved a maximum accuracy of 58%, aiming to bridge the gap between human lecturers and machines while also considering AI's psychological and emotional impacts on education. 

Implementing this AI-driven model can offer personalized learning experiences, improve the efficiency of academic assessment, and enhance the overall quality of education by closely aligning with students' individual learning needs.  

Finding a fit between CXO’s experience and AI usage in CXO decision-making: evidence from knowledge-intensive professional service firms

Authors: Poojitha Kondapaka, Sayantan Khanra, Ashish Malik, Muneza Kagzi, and Kannan Hemachandran

This paper outlines a study focusing on integrating AI in decision-making processes at the organisations' Chief Officers (CXOs) level. It emphasizes the importance of combining CXOs' professional experiences with AI's decision-making capabilities, addressing a gap in the literature regarding the co-creation value of AI and human expertise in managerial decisions. The study uses a grounded theory approach to explore the balance between AI usage and CXOs' experience in knowledge-intensive firms, including focus group discussions and in-depth interviews. Preliminary findings led to a theoretical framework on AI implementation and competitive strategy emergence from the co-creation of value by AI and CXOs' experiences. 

The model from the research can be applied in businesses by strategically integrating AI with the expertise of CXOs to enhance decision-making. Organizations can navigate complex challenges more effectively by fostering synergy between technology and human judgment, optimizing operational efficiency and strategic innovation. 

Chronic Kidney Disease Prediction Using Boosting Techniques Based on Clinical Parameters

Authors: Shahid Mohammad Ganie, Pijush Kanti Dutta Pramanik, Saurav Mallik, and Zhongming Zhao

This work introduces a novel approach to Chronic Kidney Disease prediction, highlighting the potential of ensemble learning techniques in enhancing diagnostic accuracy and early detection in the healthcare industry. The research explores the application of ensemble learning, precisely five boosting algorithms (XGBoost, CatBoost, LightGBM, AdaBoost, and Gradient Boosting), to predict Chronic Kidney Disease (CKD) using clinical data. AdaBoost performed best after preprocessing, hyperparameter tweaking, and feature selection, reaching 100% training accuracy and 98.47% testing accuracy. 

The model developed for predicting Chronic Kidney Disease (CKD) can be integrated into healthcare practices by deploying it as a diagnostic tool. It can enable healthcare providers to accurately assess the risk of CKD in patients based on clinical parameters, leading to early intervention and personalized treatment plans, thus improving patient outcomes and healthcare efficiency. 

An Ensemble Learning Approach for Diabetes Prediction using Boosting Techniques

Authors: Shahid Mohammad Ganie, Pijush Kanti Dutta Pramanik, Majid Bashir Malik, Saurav Mallik, Hong Qin 

This work investigates the effectiveness of five boosting algorithms (XGBoost, CatBoost, LightGBM, AdaBoost, and Gradient Boosting) for predicting diabetes using clinical data. The study employs various preprocessing techniques, such as imputation, Z-score, and cleaning methods, alongside data normalization, upsampling, and hyperparameter tuning to enhance prediction accuracy. The findings indicate that Gradient Boosting achieved the highest accuracy rate of 96%, outperforming other models in precision, recall, f1-score, and ROC curve evaluations. This model demonstrates significant potential for improving diabetes prediction and can be applied to other healthcare datasets with similar features. 

The model for diabetes prediction using boosting techniques can be implemented in healthcare practices by incorporating it into clinical decision support systems. This integration allows for the early detection and risk assessment of diabetes in patients, facilitating timely intervention and personalized treatment planning, thereby enhancing patient care and management efficiency.