Showing papers by "Mansaf Alam published in 2022"

Quantum-Inspired Support Vector Machines for Human Activity Recognition in Industry 4.0

Abstract: In Industry 4.0, robots and humans share roles and responsibilities. Many industries, such as manufacturing and logistics, still depend heavily on manual labor. However, by integrating sensor technology and data processing, different processes can be appropriately tracked and streamlined. Comprehensive knowledge of the occurrence, duration, and properties of related human activities is needed to draw inferences on enhancing employee performance. The increased use of numerous wearable sensors to monitor human activities has resulted in an unexpected data explosion. The pace at which this data is being generated is expected to outpace conventional technology’s ability to handle it by 2030. At present, quantum computing is the most promising market solution. This paper proposes the use of a quantum support vector machines for human activity recognition. As compared to various state-of-the-art machine learning algorithms, such as linear discriminant analysis (LDA), logistic regression, k-nearest neighbor (kNN), and conventional support vector machines (SVMs) with three different kernels, the proposed approach takes the lead by achieving an accuracy of 98% with execution time of 19 s.

Modeling the use of voice based assistant devices (VBADs): A machine learning base an exploratory study using cluster analysis and correspondence analysis

Abstract: There is a growing prevalence of Voice-Based Assistant Devices (VBADs) in our urban lives. The unique and lively user and interaction experience offered by this smart and Artificial Intelligence (AI) enabled device, not only helps to serve a variety of functions for its users but also provides them companionship to relieve the boredom. The present study attempts to explore the various describing words or phrases that the users of VBADs often use describing them, by making use of the projective word association technique. Post this using the cluster analysis the study has aimed to probe into the usage intentions of the users regarding these devices, examining if these are used for purely functional purposes or are also capable of meeting some kind of friendly relationship bond with their users. The cluster analysis was performed using machine learning and mathematical modeling. Further, the study explores a perceptual brand mapping for a few popular brands of VBADs using correspondence analysis, to understand if these brands have more points of parity in common or are they distinct across each other. The findings indicate that besides meeting the regular functional goals for the people, these devices also bond with the people by becoming their partners, friends, or members of the family. There exists a distinction across brands studied, where though all the brands satisfied their users by serving different functional purposes, however only Google Assistant and Alexa were capable of forming relational ties with their users. The brands engineering these devices can essentially base their product development and user interaction model taking the findings of this study into consideration for improving the future user experience.

IoT-Cloud Enabled Statistical Analysis and Visualization of Air Pollution Data in India

Abstract: Air pollution is a significant issue in our environment. Over the last few decades, urbanization and industrialization have accelerated in developing countries, resulting in an outsized problem of air pollution. In this paper, we have studied the air quality data gathered from IoT-enabled devices and stored onto Cloud storage infrastructure, published by the Indian government Web site. For many Indian states, we have considered the various analysis factors and tried to visualize the IoT-enabled air pollution data, and analyzed the impact of pollution on human life in our society. Furthermore, we have analyzed the air pollution on dataset based on certain conditions using the correlation and heatmap.

Integration of Convolutional Neural Networks and Recurrent Neural Networks for Foliar Disease Classification in Apple Trees

Abstract: —Automated methods intended for image classification have become increasingly popular in recent years, with applications in the agriculture field including weed identification, fruit classification, and disease detection in plants and trees. In image classification, convolutional neural networks (CNN) have already shown exceptional results but the problem with these models is that these models cannot extract some relevant image features of the input image. On the other hand, the recurrent neural network (RNN) can fully exploit the relationship among image features. In this paper, the performance of combined CNN and RNN models is evaluated by extracting relevant image features on images of diseased apple leaves. This article suggested a combination of pre-trained CNN network and LSTM, a particular type of RNN. With the use of transfer learning, the deep features were extracted from several fully connected layers of pre-trained deep models i.e. Xception, VGG16, and InceptionV3. The extracted deep features from the CNN layer and RNN layer were concatenated and fed into the fully connected layer to allow the proposed model to be more focused on finding relevant information in the input data. Finally, the class labels of apple foliar disease images are determined by the integrated model for apple foliar disease classification, experimental findings demonstrate that the proposed approach outperforms individual pre-trained models.

Envisaging Industrial Perspective Demand Response Using Machine Learning

Abstract: The utilization of Internet of Things (IoT) sensors in industry which generates the data that is used in a variety of analytics for the acquisition of valuable information is characterised as Industrial Internet of Things (IIoT). In predictive analytics, the primary aspect is typically the kind of data provided by the sensors. Various kinds of data are collected while performing predictive modelling, for, e.g., area, season, energy, cost, etc. In this paper, a case study is presented, in which a dataset is used to examine the functioning of equipment and to analyse the demand and response of that equipment. The main aim of this paper is to employ various machine learning algorithms in order to devise of predictive models in industrial IoT environment using the dataset mentioned above.

A Secure Key Agreement Framework for Cloud Computing Using ECC

Abstract: Cloud computing has the ability to share data via public channel. Digital identity is needed for clients to use facilities of cloud in cloud computing. Presently, public key and asymmetric cryptography are used for most of cloud-based system to provide safe communication. Due to its characteristics, ID-based cryptography is very important in cloud domain. Our framework has three phases like as initialization, user registration and authentication, and key agreement phase. Also, we have compared our work with other works in the same domain and also found that our work is more secure than others. In our framework, user and server communicates to each other and establish key agreement.