Praveen Gujjar Jagannath
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Praveen Gujjar Jagannath has a total of 40 co-authors. Here are the top co-authors by number of shared publications:
- Guru Prasad Marigundanahalli Siddabasappa
Graphic Era (Deemed to be University)
7 shared publications - Naveen Kumar Hullahalli Nannappa
Vidyavardhaka College of Engineering
3 shared publications Raghavendra M Devadas
Department of Information Technology, Manipal Institute of Technology Bengaluru, Manipal Academy of Higher Education (MAHE), Manipal, India
3 shared publications- Raghavendra M. Devadas
Manipal Institute of Technology Bengaluru, Manipal Academy of Higher Education, Manipal, India
3 shared publications Naveen Kumar H. N.
Vidyavardhaka College of Engineering, India
2 shared publicationsVani Hiremani
Symbiosis Institute of Technology, Symbiosis International University (Deemed), Pune, India
2 shared publications
Latest Publications
- Open Access
- 12 June 2026
- Published by Springer Nature
- p. 32-40
A Deep Ensemble Learning Model for Prediction of Attrition – Effect of Unbalanced Dataset
Deep ensemble learning has become a very popular and efficient technique for making predictions. Traditional machine learning models are often not efficient when used for making predictions of imbalanced datasets. To counter the effects of imbalanced datasets, several different oversampling and synthetic sampling methods have been proposed. The effect of such sampling methods on deep ensemble learning methods is also not well studied. This study aims to explore two basic questions: whether deep ensemble models are suitable for unbalanced data given traditional methods are not and secondly if we use the oversampling method along with deep ensemble learning, does it improve model performance? For our comparison, we choose four models. We further scope this study by using only sequential ensemble models. We choose three baseline models and two deep-learning ensemble models for our study. We find that the deep ensemble learning models are better when we consider a combination of precision and recall. Also, oversampling significantly increases the model’s ability to correctly predict the underrepresented class.
- 12 June 2026
- Published by Springer Nature
- p. 32-40
Open AccessA Deep Ensemble Learning Model for Prediction of Attrition – Effect of Unbalanced Dataset
Deep ensemble learning has become a very popular and efficient technique for making predictions. Traditional machine learning models are often not efficient when used for making predictions of imbalanced datasets. To counter the effects of imbalanced datasets, several different oversampling and synthetic sampling methods have been proposed. The effect of such sampling methods on deep ensemble learning methods is also not well studied. This study aims to explore two basic questions: whether deep ensemble models are suitable for unbalanced data given traditional methods are not and secondly if we use the oversampling method along with deep ensemble learning, does it improve model performance? For our comparison, we choose four models. We further scope this study by using only sequential ensemble models. We choose three baseline models and two deep-learning ensemble models for our study. We find that the deep ensemble learning models are better when we consider a combination of precision and recall. Also, oversampling significantly increases the model’s ability to correctly predict the underrepresented class.
- 28 January 2026
- Published by Springer Nature
- p. 471-489
Sustainable Consumer Behavior and Practical Implications at the Fashion Industry
- 28 January 2026
- Published by Springer Nature
- p. 471-489
Sustainable Consumer Behavior and Practical Implications at the Fashion Industry
- Open Access
- 19 January 2026
- Published by Prisma Publications in International Journal of Data Informatics and Intelligent Computing
- Vol. 5 (1) , 1-10
Wavelet-Based Intelligent Framework for Network Traffic Anomaly Detection in IoT Embedded Systems
The rapid growth of the Internet of Things (IoT) and embedded systems has increased the vulnerability of network infrastructures to cyber-attacks, necessitating efficient real-time anomaly detection methods. In this study, we introduce a primary IoT network traffic dataset generated through controlled simulation of normal and malicious behaviors. Both time-domain features, including packet size, inter-arrival time, protocol type, and TCP flags, and frequency-domain features, such as spectral entropy and band energy derived via wavelet transform, were extracted to capture comprehensive traffic characteristics. These features were used to train a hybrid deep learning model, the Adaptive Differential Evolution Weighted Deep Belief Network (ADE-WDBN), which combines deep hierarchical feature learning with evolutionary weight optimization to enhance detection accuracy and computational efficiency. Experimental results demonstrate that ADE-WDBN outperforms traditional machine learning models and conventional deep learning approaches, achieving an accuracy of 98.37%, precision of 97.65%, recall of 98.02%, and F1-score of 97.83%. The low variability in performance across cross-validation folds indicates the model's robustness and generalizability. This research contributes a novel IoT traffic dataset and a cost-effective, adaptive anomaly detection framework capable of detecting subtle anomalies.
- 19 January 2026
- Published by Prisma Publications in International Journal of Data Informatics and Intelligent Computing
- Vol. 5 (1) , 1-10
Open AccessWavelet-Based Intelligent Framework for Network Traffic Anomaly Detection in IoT Embedded Systems
The rapid growth of the Internet of Things (IoT) and embedded systems has increased the vulnerability of network infrastructures to cyber-attacks, necessitating efficient real-time anomaly detection methods. In this study, we introduce a primary IoT network traffic dataset generated through controlled simulation of normal and malicious behaviors. Both time-domain features, including packet size, inter-arrival time, protocol type, and TCP flags, and frequency-domain features, such as spectral entropy and band energy derived via wavelet transform, were extracted to capture comprehensive traffic characteristics. These features were used to train a hybrid deep learning model, the Adaptive Differential Evolution Weighted Deep Belief Network (ADE-WDBN), which combines deep hierarchical feature learning with evolutionary weight optimization to enhance detection accuracy and computational efficiency. Experimental results demonstrate that ADE-WDBN outperforms traditional machine learning models and conventional deep learning approaches, achieving an accuracy of 98.37%, precision of 97.65%, recall of 98.02%, and F1-score of 97.83%. The low variability in performance across cross-validation folds indicates the model's robustness and generalizability. This research contributes a novel IoT traffic dataset and a cost-effective, adaptive anomaly detection framework capable of detecting subtle anomalies.
- 21 July 2025
- Published by Taylor & Francis
- p. 88-104
Smart healthcare devices
Smart healthcare devices (SHD) are renovating the way in which healthcare is provided to patients, improving resource use, and improving overall efficiency by leveraging the cutting-edge technologies. Internet of Things (IoT) for Smart Healthcare enables smooth connectivity and exchange of data among various devices, sensors, and werables. Smart healthcare with IoT allows the healthcare professionals to monitor the patient in real time and provide the healthcare solution to patient remotely. Sensors play a vital part in collecting dynamic health data, and standardized communication protocols ensure effective data exchange between data storage and devices. Machine Learning (ML) Technologies also enhance the smart healthcare systems by providing various algorithms for predictive analytics, diagnostics, and personalized treatment plans. Cloud and fog computing technologies provide the secure and expendable infrastructure for processing and storing patient data. It also supports real-time analytics. This work also illustrates the challenges related to data privacy, security, and regulatory compliance. Provide the potential healthcare delivery to improve patient outcomes and enhance the overall quality of care.
- 21 July 2025
- Published by Taylor & Francis
- p. 88-104
Smart healthcare devices
Smart healthcare devices (SHD) are renovating the way in which healthcare is provided to patients, improving resource use, and improving overall efficiency by leveraging the cutting-edge technologies. Internet of Things (IoT) for Smart Healthcare enables smooth connectivity and exchange of data among various devices, sensors, and werables. Smart healthcare with IoT allows the healthcare professionals to monitor the patient in real time and provide the healthcare solution to patient remotely. Sensors play a vital part in collecting dynamic health data, and standardized communication protocols ensure effective data exchange between data storage and devices. Machine Learning (ML) Technologies also enhance the smart healthcare systems by providing various algorithms for predictive analytics, diagnostics, and personalized treatment plans. Cloud and fog computing technologies provide the secure and expendable infrastructure for processing and storing patient data. It also supports real-time analytics. This work also illustrates the challenges related to data privacy, security, and regulatory compliance. Provide the potential healthcare delivery to improve patient outcomes and enhance the overall quality of care.
- 1 July 2025
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 84 (37) , 46405-46433
Robustness of facial expression recognition systems in the presence of mild and uniform blur
- 1 July 2025
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 84 (37) , 46405-46433
Robustness of facial expression recognition systems in the presence of mild and uniform blur
- Open Access
- 3 March 2025
- Published by Indian Society for Education and Environment in Indian Journal of Science and Technology
- Vol. 18 (8) , 580-585
Agri Friendly Conversational AI Chatbot Using Open Source Framework
Objectives: To develop an agriculturally focused conversational AI chatbot using the Rasa Open Source framework and to create a user-friendly interface tailored to the needs of the farming community. The chatbot aims to provide accessible and cost-effective solutions by offering real-time assistance, information, and support to farmers. Method: The methodology involves leveraging Rasa’s advanced natural language understanding (NLU) capabilities to enable accurate intent recognition and entity extraction. Agricultural data, including FAQs, crop details, pest control methods, and weather information, are collected and used to train machine learning models. The chatbot is refined through iterative testing and feedback from users in the agricultural domain. Deployment is planned on mobile and web platforms to ensure wide accessibility. Findings: The findings indicate that Rasa Open Source is a robust framework that allows for the creation of scalable, context-aware, and customizable conversational agents. The integration of advanced NLP techniques enhances the chatbot’s ability to comprehend and respond to user inputs effectively. Novelty: The novelty of this research lies in its application of open-source conversational AI technology to agriculture, addressing the specific challenges faced by farmers while maintaining cost efficiency. This contributes significantly to the growing use of AI in agriculture, providing a practical tool for decision-making and problem-solving. Unlike traditional tools, this chatbot integrates natural language understanding and dialogue management to deliver real-time assistance on critical topics such as weather updates, crop information, pest control, and market insights. Keywords: Agriculture Friendly chatbot, AI chatbot, Rule Based chatbot, RASA Transfer Learning
- 3 March 2025
- Published by Indian Society for Education and Environment in Indian Journal of Science and Technology
- Vol. 18 (8) , 580-585
Open AccessAgri Friendly Conversational AI Chatbot Using Open Source Framework
Objectives: To develop an agriculturally focused conversational AI chatbot using the Rasa Open Source framework and to create a user-friendly interface tailored to the needs of the farming community. The chatbot aims to provide accessible and cost-effective solutions by offering real-time assistance, information, and support to farmers. Method: The methodology involves leveraging Rasa’s advanced natural language understanding (NLU) capabilities to enable accurate intent recognition and entity extraction. Agricultural data, including FAQs, crop details, pest control methods, and weather information, are collected and used to train machine learning models. The chatbot is refined through iterative testing and feedback from users in the agricultural domain. Deployment is planned on mobile and web platforms to ensure wide accessibility. Findings: The findings indicate that Rasa Open Source is a robust framework that allows for the creation of scalable, context-aware, and customizable conversational agents. The integration of advanced NLP techniques enhances the chatbot’s ability to comprehend and respond to user inputs effectively. Novelty: The novelty of this research lies in its application of open-source conversational AI technology to agriculture, addressing the specific challenges faced by farmers while maintaining cost efficiency. This contributes significantly to the growing use of AI in agriculture, providing a practical tool for decision-making and problem-solving. Unlike traditional tools, this chatbot integrates natural language understanding and dialogue management to deliver real-time assistance on critical topics such as weather updates, crop information, pest control, and market insights. Keywords: Agriculture Friendly chatbot, AI chatbot, Rule Based chatbot, RASA Transfer Learning
- 27 September 2024
- Published by IGI Global Scientific Publishing
- p. 107-128
Unveiling the Potential of Generative Approaches in AI-Infused Web Development for Design, Testing, and Maintenance
The digital landscape is witnessing the continuous evolution of our lived experiences, in which the cross-cutting domain of Generative Artificial Intelligence (AI) and web engineering models is emerging as a pivotal source of transformative innovation. This article seeks to identify the potential new opportunities that generative AI could explore since it heralds in a radical revolution in the traditional paradigms associated with web development. The studied problems addressed in this paper include historical web system construction issues, design complexities, and maintainability problems. The research concludes by presenting a roadmap for embracing this paradigm shift, contributing to the ongoing discourse on the intersection of AI and web engineering. The study is exploratory at the outset, trying to gain insights into the integration of generative AI into web development. This research aims to explore the challenges, benefits, and potential applications of generative AI in several phases of web development, including design, testing, and maintenance.
- 27 September 2024
- Published by IGI Global Scientific Publishing
- p. 107-128
Unveiling the Potential of Generative Approaches in AI-Infused Web Development for Design, Testing, and Maintenance
The digital landscape is witnessing the continuous evolution of our lived experiences, in which the cross-cutting domain of Generative Artificial Intelligence (AI) and web engineering models is emerging as a pivotal source of transformative innovation. This article seeks to identify the potential new opportunities that generative AI could explore since it heralds in a radical revolution in the traditional paradigms associated with web development. The studied problems addressed in this paper include historical web system construction issues, design complexities, and maintainability problems. The research concludes by presenting a roadmap for embracing this paradigm shift, contributing to the ongoing discourse on the intersection of AI and web engineering. The study is exploratory at the outset, trying to gain insights into the integration of generative AI into web development. This research aims to explore the challenges, benefits, and potential applications of generative AI in several phases of web development, including design, testing, and maintenance.
- 24 August 2024
- Published by Springer Nature
- p. 283-303
Innovative Fusion: Attention-Augmented Support Vector Machines for Superior Text Classification for Social Marketing
- 24 August 2024
- Published by Springer Nature
- p. 283-303
Innovative Fusion: Attention-Augmented Support Vector Machines for Superior Text Classification for Social Marketing
- 6 August 2024
- Published by IGI Global Scientific Publishing
- p. 127-152
Aiding in Prognosis of Fibromyalgia Syndrome Utilizing MCDM TOPSIS Method in Fusion With Fuzzy Logic
The study addresses the complexity of Fibromyalgia syndrome (FMS) by integrating multi-criteria decision-making (MCDM) TOPSIS with fuzzy logic for accurate prognosis. A comprehensive set of criteria, including clinical, psychological, and physiological factors, is employed. Fuzzy logic models uncertainties and subjective expert opinions. The method systematically evaluates and ranks FMS prognoses, contributing to transparency in decision-making. Incorporating sensitivity analysis with small variations enhances reliability. Considering symptoms, patients, and expert ratings, the study identifies patient 5 as more likely to have FMS (closeness coefficient 0.881), while patient 1 is less prone (closeness coefficient 0.088).
- 6 August 2024
- Published by IGI Global Scientific Publishing
- p. 127-152
Aiding in Prognosis of Fibromyalgia Syndrome Utilizing MCDM TOPSIS Method in Fusion With Fuzzy Logic
The study addresses the complexity of Fibromyalgia syndrome (FMS) by integrating multi-criteria decision-making (MCDM) TOPSIS with fuzzy logic for accurate prognosis. A comprehensive set of criteria, including clinical, psychological, and physiological factors, is employed. Fuzzy logic models uncertainties and subjective expert opinions. The method systematically evaluates and ranks FMS prognoses, contributing to transparency in decision-making. Incorporating sensitivity analysis with small variations enhances reliability. Considering symptoms, patients, and expert ratings, the study identifies patient 5 as more likely to have FMS (closeness coefficient 0.881), while patient 1 is less prone (closeness coefficient 0.088).
- 6 August 2024
- Published by IGI Global Scientific Publishing
- p. 74-88
Emerging Implications of Metaverse in the Healthcare Domain
The Metaverse, often referred to as the immersive internet, is widely considered the next significant technological disruption on the horizon, with the potential to reshape clinician-patient interactions, enhance the patient experience, transform innovation and research and development processes. The Metaverse is currently in its developmental phase, and the establishment of a definitive framework is an ongoing endeavor. In recent years, the concept of the Metaverse has gained substantial traction and has evolved into a multifaceted virtual universe with limitless possibilities. This chapter provides a glimpse into the evolving landscape of healthcare, where the Metaverse's immersive and interconnected experiences have the power to revolutionize how we perceive, access, and deliver healthcare services. From virtual clinics and medical simulations to AI-assisted diagnostics, this chapter explores the multifaceted ways in which the Metaverse is reshaping healthcare and creating new opportunities for improved patient outcomes, education, and research.
- 6 August 2024
- Published by IGI Global Scientific Publishing
- p. 74-88
Emerging Implications of Metaverse in the Healthcare Domain
The Metaverse, often referred to as the immersive internet, is widely considered the next significant technological disruption on the horizon, with the potential to reshape clinician-patient interactions, enhance the patient experience, transform innovation and research and development processes. The Metaverse is currently in its developmental phase, and the establishment of a definitive framework is an ongoing endeavor. In recent years, the concept of the Metaverse has gained substantial traction and has evolved into a multifaceted virtual universe with limitless possibilities. This chapter provides a glimpse into the evolving landscape of healthcare, where the Metaverse's immersive and interconnected experiences have the power to revolutionize how we perceive, access, and deliver healthcare services. From virtual clinics and medical simulations to AI-assisted diagnostics, this chapter explores the multifaceted ways in which the Metaverse is reshaping healthcare and creating new opportunities for improved patient outcomes, education, and research.