Karthik Vasu
Research Subjects
Karthik Vasu has a total of 5 co-authors. Here are the top co-authors by number of shared publications:
- Amalraj Irudayasamy
College of Computing and Information Sciences, University of Technology and Applied Sciences-Nizwa, Oman
1 shared publication - Anitha Premkumar
Presidency University
1 shared publication - Natesh Mahadev
Department of Computer Science and Engineering, Vidyavardhaka College of Engineering, India
1 shared publication - Rajesh Natarajan
Department of Computing and Information Sciences, College of Computing and Information Sciences, University of Technology and Applied Sciences, Shinas, Oman
1 shared publication - Sujatha Krishna
Department of Computing and Information Sciences, College of Computing and Information Sciences, University of Technology and Applied Sciences, Oman
1 shared publication
Latest Publications
- Open Access
- 1 January 2026
- Published by Wiley in Applied Computational Intelligence and Soft Computing
- Vol. 2026 (1)
AI‐Based Air Pollution Forecasting Using SAE and Metaheuristic‐Optimized ANN
Artificial intelligence (AI) enables data‐driven modeling and forecasting for smart city environmental monitoring. Government agencies and the public are becoming more concerned about air quality, which impacts several aspects of the human environment and development. Through careful analysis of environmental and pollution data, the research identified key factors that influence air pollution monitoring. Traditional air pollution forecasting methods often show limited predictive accuracy due to simplified modeling assumptions and the absence of heuristic optimization strategies for model parameter tuning. This research aimed to develop intelligent ecosystem models for smart cities using a battle royal optimized multilevel artificial neural network (BRO‐MLANN). The Global Air Pollution dataset from Kaggle, which includes air quality observations from multiple towns, is used to assess the suggested model. The goal of this endeavor is to improve air quality prediction in smart cities by using an intelligent ecosystem model based on BRO‐MLANN. The results show how well the framework supports air quality monitoring and decision‐making in smart city environments, even if the dataset encompasses a variety of urban areas and does not represent all worldwide regions. Prior to model training, Z ‐score normalization was used to standardize the data. The stacked autoencoder (SAE) used a feature extraction technique to predict air quality to plan intelligent environment design for smart cities. In developing intelligent ecosystem methods for smart cities, the BRO‐MLANN is utilized. The results show that the BRO‐MLANN method achieves higher performance scores for prediction rate, MAE, RMSE, MAPE, and accuracy compared to other competing techniques.
- 1 January 2026
- Published by Wiley in Applied Computational Intelligence and Soft Computing
- Vol. 2026 (1)
Open AccessAI‐Based Air Pollution Forecasting Using SAE and Metaheuristic‐Optimized ANN
Artificial intelligence (AI) enables data‐driven modeling and forecasting for smart city environmental monitoring. Government agencies and the public are becoming more concerned about air quality, which impacts several aspects of the human environment and development. Through careful analysis of environmental and pollution data, the research identified key factors that influence air pollution monitoring. Traditional air pollution forecasting methods often show limited predictive accuracy due to simplified modeling assumptions and the absence of heuristic optimization strategies for model parameter tuning. This research aimed to develop intelligent ecosystem models for smart cities using a battle royal optimized multilevel artificial neural network (BRO‐MLANN). The Global Air Pollution dataset from Kaggle, which includes air quality observations from multiple towns, is used to assess the suggested model. The goal of this endeavor is to improve air quality prediction in smart cities by using an intelligent ecosystem model based on BRO‐MLANN. The results show how well the framework supports air quality monitoring and decision‐making in smart city environments, even if the dataset encompasses a variety of urban areas and does not represent all worldwide regions. Prior to model training, Z ‐score normalization was used to standardize the data. The stacked autoencoder (SAE) used a feature extraction technique to predict air quality to plan intelligent environment design for smart cities. In developing intelligent ecosystem methods for smart cities, the BRO‐MLANN is utilized. The results show that the BRO‐MLANN method achieves higher performance scores for prediction rate, MAE, RMSE, MAPE, and accuracy compared to other competing techniques.