DKD. Arul Kumar

D. Arul Kumar

*This is an algorithmically generated author record based on publicly available information.

D. Arul Kumar has a total of 11 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    IoT-Driven Adaptive Controlled-Traffic Farming Framework with Multi-Objective Optimization for Sustainable Groundnut Cultivation and Soil Compaction Reduction in Andhra Pradesh

    by R. Mohanraj, Gokula Krishnan Vasudevan, Gummadi Reshma, J. Lydia Pancy, D. Arul Kumar, R. Aiyshwariya Devi

    Groundnut is an important crop in the agriculture economy of Andhra Pradesh and has high requirement of mechanical operations which may lead to soil compaction and hence its impact on soil structure, root zone and crop productivity. Australian Controlled-Traffic Farming (CTF) would decrease the soil degradation caused by machinery, but the heterogeneous field condition, fragmented land ownership, and development of intelligent decision-support would hinder the adoption of CTF in Indian agriculture. In this study, an Internet of Things (IoT) driven Adaptive Multi-Objective Controlled-Traffic Optimisation (AMO-CTO) framework is proposed to develop a digital agricultural twin for sustainable groundnut farming in the publicly available Area, Production and Yield (APY) data and using real-time observations from IoT sensors. The framework incorporates adaptive traffic-lane optimisation, dynamic soil-compaction prediction, groundnut-yield estimation and multi-objective resource optimisation. The experimental evaluation results show that the proposed framework has a prediction accuracy of 95.82% and an R2 value of 0.987. The AMO-CTO framework under the evaluation situations of this study has resulted in yield improvement of groundnut by 29.7% and reduction in soil compaction by 32.8%, irrigation-water by 26.4% and fuel by 23.9%. These enhancements are dependent on the particular reported experimental procedure and should not be considered to be widely applicable field-level results.

  • A Modified Generative Adversarial Network-based Drug Discovery Model for Accelerating Biomolecular Interaction Predictions

    by V. Gokula Krishnan, Shanker M.C, D. Arul Kumar, S. Kaviarasan, G. Sandra Karunya, B. Prathusha Laxmi

    The drug discovery model for accelerating biomolecular interaction predictions employs sophisticated deep learning methodologies to recognize and foresee drug-target molecule interactions. This model makes use of large-scale databases such as GDSC and CCLE to increase the precision of the interaction mapping. Drug development is accelerated due to the time and cost savings. This paper proposes a modified One-Dimensional Convolutional Wasserstein Generative Adversarial Network (1D-CWGAN) for improving drug response prediction by tackling data imbalance in biomolecular datasets. The framework that is put forward mixes data augmentation using CWGAN with CNN and BiSRU based predictive models, alongside MGS-ROA for effective hyper parameter optimization. Employing sequential feature learning and stabilized adversarial training, the model produces top-notch synthetic samples that boost prediction stability. Tests carried out on GDSC and CCLE datasets show that the proposed CWGAN-BiSRU model is superior in performance, attaining RMSE values of 0.023 and 0.024, respectively, and higher correlation and determination coefficients than traditional machine learning, deep learning, and GAN-based methods. The findings reveal an improved ability to generalize and the efficiency of the proposed method for prediction of biomolecular interactions in drug discovery applications.

  • Open Access

    AOA based Masked Region-CNN model for Detection of Parking Space in IoT Environment

    by Sri Vijaya K, Gokula Krishnan V, Arul Kumar D, Prathusha Laxmi B, Yasaswi B

    Uneven illumination has a significant impact on vision-based automatic parking systems, making it impossible to conduct a correct assessment of parking places in the presence of complicated picture data. In to address this issue, this work provides a deep learning-based system for visual recognition of parking spaces and picture processing. Artificial intelligence (AI) approaches can be used to identify a less expensive and easier-to-implement solution to the parking spot identification challenge, especially since the discipline of deep learning is reshaping the world. Using deep learning techniques, this study offers a dynamic, straightforward, and cost-effective algorithm for the detection of parking spots. In order to determine which parking spots are available and which are occupied, this method employs a Masked Region Based Convolutional Neural Network (MR-CNN) and the intersection over union approach. Cars in the training dataset were spaced more apart than those actually seen, which increased the accuracy of the identification between cars and parking spots. The AOA mechanism enhances the model's ability to focus on relevant regions within an image, improving accuracy in detecting parking spaces. This leads to precise identification of parking slots, reducing false positives and negatives. The sequence and quantity of parking spots, as well as the capacity to predict empty spots, were tested in a case study and found to be accurate. In the experimental results as the AOA based MR-CNN model stretched the accuracy as 98.50 and the recall value as 40.59 then the precision as 96.34 F1-measure as 57.95 correspondingly.