scispace - formally typeset
Search or ask a question
Author

Meenakshisundaram Iyapparaja

Bio: Meenakshisundaram Iyapparaja is an academic researcher from VIT University. The author has contributed to research in topics: Search algorithm & Feature selection. The author has an hindex of 1, co-authored 3 publications receiving 4 citations.

Papers
More filters
Journal ArticleDOI
TL;DR: This research has extracted the Big Data SIoT using the well-known model named MapReduce framework and the implementation of the proposed GA-EHO is done by using some machine learning classifiers for classifying the data and the efficiency is predicted for the proposed work.
Abstract: Several novel applications and services of networking for the IoT are supported by the Social Internet of Things (SIoT) in a more productive and powerful way. SIoTs are the recent hot topics rather than other extensions of IoTs. In this research, the authors have extracted the Big Data SIoT using the well-known model named MapReduce framework. Moreover, the unwanted data and noise from the database are reduced using the Gabor filter, and the big databases are mapped and reduced using the Hadoop MapReduce (HMR) technique for improving the efficiency of the proposed GA-EHO. Furthermore, the feature selection using GA-EHO is processed on the filtered dataset. The implementation of the proposed system is done by using some machine learning classifiers for classifying the data and the efficiency is predicted for the proposed work. From the simulation results, the specificity, maximum accuracy, and sensitivity of the proposed GA-EHO are produced about 87.88%, 99.1%, and 81%. Also, the results are compared with other existing techniques.

5 citations

Journal ArticleDOI
TL;DR: The prominent features of both wavelet model and GSA model are obtained and are shared with the charged system search algorithm to minimise the total cost experienced for location area optimisation in mobile wireless communication networks (MWCN).
Abstract: Location area optimisation is used to diminish the location update cost and paging cost in mobile wireless communication networks. Retaining heuristic optimisation technique, helps to diminish the location and paging cost, the problem occurs in this technique is combinational optimisation in nature. Handiness of mobile users growing day by day and many users will be allocated to various mobile subscribers and thus forecasting the ideal area is always a big job. Charged system search algorithm (CSSA) is employed to overwhelm the local and global minima that happened often during the peers of the run process. The variants introduced into the CSSA include the wavelet models and the gravitational search algorithm (GSA) models. The prominent features of both wavelet model and GSA model are obtained and are shared with the charged system search algorithm to minimise the total cost experienced for location area optimisation in mobile wireless communication networks (MWCN).

1 citations

Journal ArticleDOI
TL;DR: The machine learning techniques that were used in previous works were analysed, the performance comparison of various classifiers on different datasets is shown and SVM has more than 90% of accuracy when compared with other algorithms.
Abstract: The rapid development of technologies in today's world has become interesting that made millions of people to utilise the major advantages in it. Two main technologies that were emerging in modern society are big data and the social internet of things. Several researchers have studied and developed a major concept of using big data with SIoT and the security development of maintain a large amount of data. In this paper, deep survey regarding the concepts behind the big data analytics with the social internet of things (SIoT) was studied and analysed. Furthermore, the machine learning techniques that were used in previous works were analysed and comparisons of various methods are discussed. The performance comparison of various classifiers on different datasets is shown and SVM has more than 90% of accuracy when compared with other algorithms. KNN has 64% of accuracy which is lowest of any classifier than NB and NN.

1 citations


Cited by
More filters
Journal ArticleDOI
TL;DR: In this article, the authors discuss the role of IoT in social relationships detection and management, the problem of social relationships explosion in IoT and review the proposed solutions using ASI, including social-oriented machine-learning and deep-learning techniques.
Abstract: With the recent advances of the Internet of Things, and the increasing accessibility of ubiquitous computing resources and mobile devices, the prevalence of rich media contents, and the ensuing social, economic, and cultural changes, computing technology and applications have evolved quickly over the past decade. They now go beyond personal computing, facilitating collaboration and social interactions in general, causing a quick proliferation of social relationships among IoT entities. The increasing number of these relationships and their heterogeneous social features have led to computing and communication bottlenecks that prevent the IoT network from taking advantage of these relationships to improve the offered services and customize the delivered content, known as relationship explosion. On the other hand, the quick advances in artificial intelligence applications in social computing have led to the emerging of a promising research field known as Artificial Social Intelligence (ASI) that has the potential to tackle the social relationship explosion problem. This paper discusses the role of IoT in social relationships detection and management, the problem of social relationships explosion in IoT and reviews the proposed solutions using ASI, including social-oriented machine-learning and deep-learning techniques.

20 citations

Journal ArticleDOI
TL;DR: In this paper, the authors proposed a Stratified Systematic Sampling Extension (SSE) approach for risk analysis in big data mining using a single machine execution by clustering methodology.
Abstract: Risk analysis is one of the most essential business activities because it discovers unknown risks such as financial risk, recovery risk, investment risk, operational risk, credit risk, debit risk, and so on. Clustering is a data mining technique that uses data behavior and nature to discover unexpected risks in business data. In a big data setup, clustering algorithms encounter execution time and cluster quality-related challenges due to the primary attribute of big data. This study suggests a Stratified Systematic Sampling Extension (SSE) approach for risk analysis in big data mining using a single machine execution by clustering methodology. Sampling is a data reduction technique that saves computation time and improves cluster quality, scalability, and speed of the clustering algorithm. The proposed sampling plan first formulates the stratum by selecting the minimum variance dimension and then selects samples from each stratum using random linear systematic sampling. The clustering algorithm produces robust clusters in terms of risk and non-risk group with the help of sample data and extends the sample-based clustering results to final clustering results utilizing Euclidean distance. The performance of the SSE-based clustering algorithm has been compared to existing K-means and K-means ++ algorithms using Davies Bouldin score, Silhouette coefficient, Scattering Density between clusters Validity, Scattering Distance Validity and CPU time validation metrics on financial risk datasets. The experimental results demonstrate that the SSE-based clustering algorithm achieved better clustering objectives in terms of cluster compaction, separation, density, and variance while minimizing iterations, distance computation, data comparison, and computational time. The statistical analysis reveals that the proposed sampling plan attained statistical significance by employing the Friedman test.

9 citations

Journal ArticleDOI
TL;DR: In this article , the Integrated Transfer Learning-based Convolutional Neural Network (ITL-CNN) model was proposed to improve the classification accuracy for the detection of polycystic ovarian syndrome using ultrasound images.
Abstract: In recent years, Polycystic Ovary Syndrome (PCOS) becomes one of the most prominent research areas, where several researchers are concentrating to improve the accuracy of PCOS classification. It is much difficult to find the presence of PCOS in women with traditional techniques and various researchers are dealt with the problem that affects the accuracy in detecting such symptom. In this paper, we have proposed Integrated Transfer Learning-based Convolutional Neural Network (ITL-CNN) model to improve the classification accuracy for the detection of PCOS using ultrasound images. In this proposed model, we have used active contour with modified Otsu method and Multifactor Dimension Reduction-based GIST feature extractor for improving the performance of the ITL-CNN model. The performance of the proposed model is analyzed using various performance metrics such as accuracy, sensitivity, precision, recall, and F1 score. Furthermore, the results show that the proposed ITL-CNN model outperforms by achieving 98.9% of accuracy when compared with other existing techniques such as Convolutional Neural Network (CNN), Artificial Neural Network (ANN), Support Vector Machine (SVM), and Gaussian Naïve Bayes (NB).

2 citations

Proceedings ArticleDOI
20 Nov 2022
TL;DR: In this article , the authors present a brief assessment of the current approaches to pest identification and categorization using deep learning and machine learning models, and suggest a structure for the identification and classification of pests by means of the transfer learning models Pla-Net and GoogleNet.
Abstract: Agriculture is the foundation of many civilizations in addition to being the breath of survival among many living things. One of the most crucial tasks on this farming is pest control. The yield of agricultural crops is significantly impacted by insect infestations. Accurate and quick identification of parasitic insects in the earlier phases owing to deep learning algorithms advancements may save both micro- and macro significant disruption. The proposal's objective is to identify insect pests that have an impact on crop output in agriculture. This paper presents a brief assessment of the current approaches to pest identification and categorization using deep learning and machine learning models. Furthermore, it suggests a structure for the identification and classification of pests by means of the transfer learning models Pla-Net and GoogleNet. Many cattle ranchers are now earning a sizable income owing to the automation of a virus prediction model. This can improve the nation's food production by using pre-trained models and deep learning approaches. The system is evaluated using a sample dataset, and the outcomes are tested to demonstrate the effectiveness of the suggested strategy. Therefore as adopt to this work, this intends to create an efficient automated early pest detection and classification system that can help farmers more effectively and revolutionize the way farming is practiced.
Journal ArticleDOI
TL;DR: In this paper, a multi-objective analysis and comparison of the most common location update strategies jointly with most common paging procedures is presented, with the exception of blanket paging that has shown to be the most inefficient in the whole objective space.