International Journal of Next-Generation Computing

Machine Learning-assisted Distance Based Residual Energy Aware Clustering Algorithm for Energy Efficient Information Dissemination in Urban VANETs

Amit Choksi, Mehul Shah — Tue, 14 May 2024 00:00:00 +0530

A Vehicular Ad-hoc Network (VANET) is an essential component of intelligent transportation systems in the building of smart cities. A VANET is a self-configure high mobile and dynamic potential wireless ad-hoc network that joins all vehicle nodes in a smart city to provide in-vehicle infotainment services to city administrators and residents. In the smart city, the On-board Unit (OBU) of each vehicle has multiple onboard sensors that are used for data collection from the surrounding environment. One of the main issues in VANET is energy efficiency and balance because the small onboard sensors can’t be quickly recharged once installed on On-board Units (OBUs). Moreover, conserving energy stands out as a crucial challenge in VANET which is primarily contingent on the selection of Cluster Heads (CH) and the adopted packet routing strategy. To address this issue, this paper proposes distance and energy-aware clustering algorithms named SOMNNDP, which use a Self-Organizing Map Neural Network (SOMNN) machine learning technique to perform faster multi-hop data dissemination. Individual Euclidean distances and residual node energy are considered as mobility parameters throughout the cluster routing process to improve and balance the energy consumption among the participating vehicle nodes. This maximizes the lifetime of VANET by ensuring that all intermediate vehicle nodes use energy at approximately the same rate. Simulation findings demonstrate that SOMNNDP improves Quality of Service (QoS) better and consumes 17% and 14% less energy during cluster routing than distance and energy-aware variation of K-Means (KM) and Fuzzy C-Means (FCM) called KMDP and FCMDP respectively.

Alzheimer’s Disease Classification using Feature Enhanced Deep Convolutional Neural Networks

R Sreemathy, Danish Khan, Kisley Chandra, Tejas Bora, Soumya Khurana — Tue, 14 May 2024 00:00:00 +0530

Neurodegenerative disorders are one of the most insidious disorders that affect millions around the world. Presently, these disorders do not have any remedy, however, if detected at an early stage, therapy can prevent further degeneration. This study aims to detect the early onset of one such neurodegenerative disorder called Alzheimer’s Disease, which is the most prevalent neurological disorder using the proposed Convolutional Neural Network (CNN). These MRI scans are pre-processed by applying various filters, namely, High-Pass Filter, Contrast Stretching, Sharpening Filter, and Anisotropic Diffusion Filter to enhance the Biomarkers in MRI images. A total of 21 models are proposed using different preprocessing and enhancement techniques on transverse and sagittal MRI images. The comparative analysis of the proposed five-layer Convolutional Neural Network (CNN) model with Alex Net is presented. The proposed CNN model outperforms AlexNet and achieves an accuracy of 99.40%, with a precision of 0.988, and recall of 1.00, by using an edge enhanced, contrast stretched, anisotropic diffusion filter. The proposed method may be used to implement automated diagnosis of neurodegenerative disorders.

High Utility Itemset Extraction using PSO with Online Control Parameter Calibration

LOGESWARAN K, SURESH S, SAVITHA S, ANANDAMURUGAN S — Tue, 14 May 2024 00:00:00 +0530

This study investigates the use of evolutionary computation for mining high-value patterns from benchmark datasets. The approach employs a fitness function to assess the usefulness of each pattern. However, the effectiveness of evolutionary algorithms heavily relies on the chosen strategy parameters during execution. Conventional methods set these parameters arbitrarily, often leading to suboptimal solutions. To address this limitation, the research proposes a method for dynamically adjusting strategy parameters using temporal difference approaches, a machine learning technique called Reinforcement Learning (RL). Specifically, the proposed IPSO RLON algorithm utilizes SARSA learning to intelligently adapt the Crossover Rate and Mutation Rate within the Practical Swarm Optimization Algorithm. This allows IPSO RLON to effectively mine high-utility itemsets from the data.The key benefit of IPSO RLON lies in its adaptive control parameters. This enables it to discover optimal high-utility itemsets when applied to various benchmark datasets. To assess its performance, IPSO RLON is compared to existing approaches like HUPEUMU-GRAM, HUIM-BPSO, IGA RLOFF, and IPSO RLOFF using metrics like execution time, convergence speed, and the percentage of high-utility itemsets mined. From the evaluation it is observed that the proposed IPSO RLON perfroms better than the other methodology.

Deep Learning based Semantic Segmentation for Buildings Detection from Remote Sensing Images

Miral J Patel, Hasmukh P Koringa — Tue, 14 May 2024 00:00:00 +0530

Building extraction from remote sensing images is the process of automatically identifying and extracting the boundaries of buildings from high-resolution aerial or satellite images. The extracted building footprints can be used for a variety of applications, such as urban planning, disaster management, city development, land management, environmental monitoring, and 3D modeling. The results of building extraction from remote sensing images depend on several factors, such as the quality and resolution of the image and the choice of algorithm.The process of building extraction from remote sensing images typically involves a series of steps, including image pre-processing, feature extraction, and classification. Building extraction from remote sensing images can be challenging due to factors such as varying building sizes and shapes, shadows, and occlusions. However, recent advances in deep learning and computer vision techniques have led to significant improvements in the accuracy and efficiency of building extraction methods. This research presents a deep learning semantic segmentation architecture-based model for developing building detection from high resolution remote sensing images. The open-source Massachusetts dataset is used to train the suggested UNet architecture. The model is optimized using the RMSProp algorithm with a learning rate of 0.0001 for 100 epochs. After 1.52 hours of training on Google Colab the model achieved an 83.55% F1 score, which indicates strong precision and recall.

Integrating Smartphone Sensor Technology to Enhance Fine Motor and Working Memory Skills in Pediatric Obesity: A Gamified Approach

Sudipta Saha, Saikat Basu, Koushik Majumder, Sourav Das — Tue, 14 May 2024 00:00:00 +0530

Childhood obesity remains a pervasive global challenge, often accompanied by deficits in working memory and fine motor skills among affected children. These deficits detrimentally impact academic performance. Despite limited evidence, home-based interventions targeting both fine motor skills and working memory remain underexplored. Leveraging game-based approaches holds promise in behavior modification, self-management of chronic conditions, therapy adherence, and patient monitoring. In this study, a novel smartphone-based game was meticulously developed to target the enhancement of working memory and fine motor skills in a cohort of thirty-two obese or overweight children. Over two weeks, participants engaged in regular gameplay sessions within the comfort of their homes. Pretest and post-test assessments yielded compelling evidence of significant improvements, with statistical significance established at a robust 95% confidence level. Notably, participants exhibited a progressive trend of improvement in their gameplay performance. Recognizing the profound impact of academic achievement on future socioeconomic trajectories, regardless of weight management outcomes, the importance of bolstering cognitive skills cannot be overstated. This innovative intervention offers a pragmatic and cost-effective solution to empower children to cultivate essential cognitive abilities within their home environment. By fostering the development of working memory and fine motor skills, this intervention holds promise in facilitating improved academic performance and, consequently, enhancing long-term prospects for these children.