Supervised Learning based E-mail/ SMS Spam Classifier Satendra Kumar, Raj Kumar, Ashish Saini Recent Advances in Computer Science and Communications, 2025 Background: One of the challenging problems facing the modern Internet is spam, which can annoy individual customers and wreak financial havoc on businesses. Spam communications target customers without their permission and clog their mailboxes. They consume more time and organizational resources when checking for and deleting spam. Even though most web users openly dislike spam, enough are willing to accept lucrative deals that spam remains a real problem. While most web users are well aware of their hatred of spam, the fact that enough of them still click on commercial offers means spammers can still make money from them. While most customers know what to do, they need clear instructions on avoiding and deleting spam. No matter what you do to eliminate spam, you won't succeed. Filtering is the most straightforward and practical technique in spam-blocking strategies. Method: We present procedures for identifying emails as spam or ham based on text classification. Different methods of e-mail organization preprocessing are interrelated, for example, applying stop word exclusion, stemming, including reduction and highlight selection strategies to extract buzzwords from each quality, and finally, using unique classifiers to Quarantine messages as spam or ham. Results: The Nave Bayes classifier is a good choice. Some classifiers, such as Simple Logistic and Adaboost, perform well. However, the Support Vector Machine Classifier (SVC) outperforms it. Therefore, the SVC makes decisions based on each case's comparisons and perspectives. Conclusion: Many spam separation studies have focused on recent classifier-related challenges. Machine Learning (ML) for spam detection is an important area of modern research. Today, spam detection using ML is an important area of research. Examine the adequacy of the proposed work and recognize the application of multiple learning estimates to extract spam from emails. Similarly, estimates have also been scrutinized.
Development of an analytical model of drain current for junction less GAA MOSFET including source/drain resistance Applied Data Science and Smart Systems, 2024
Adapted rank order clustering-based test case prioritization for software product line testing Satendra Kumar, Raj Kumar, Ashish Saini, Monika Rani International Journal of Advanced Intelligence Paradigms, 2024 Software product line testing (SPLT) is a strenuous task due to the explosion of derivable products. It is infeasible to test all the products of a software product line (SPL), so several contributions have been presented to overcome this issue by reducing the number of products. However, not much consideration has been given to the test order of the products. Test case prioritization (TCP) technique arranges the test cases in a sequence to meet a specific performance goal. TCP is required to increase the effectiveness and efficiency of fault detection. In SPL, TCP technique arranges the configurations of products in order to be tested. Adapted rank order clustering (AROC)-based TCP approach is proposed for SPLT. Our AROC method utilises Binary Weight and Decimal Weight to arrange the products of an SPL. The results of the rigorous experimentation using AROC-based TCP approach are better than the random order and similarity-based order in terms of fault detection rate.
An Energy-efficient Ferry-based Routing Algorithm for Scattered FANET Networks Ashish Saini, Sanjeev Kumar, Amrita Kumari, Satender Kumar, Amit Kumar, Amit Kumar Proceedings of International Conference on Computational Intelligence and Sustainable Engineering Solution Cises 2023, 2023 Flying ad hoc networks (FANETs), have a wide range of applications in both civilian and military settings. Although, prevailing routing protocols are effective in well-connected networks, they struggle in sparse networks. To address this issue in FANETs, a new geographic routing method, utilizing a delay-tolerant routing protocol, is proposed in this paper. The proposed method combines position routing and delay tolerant network (DTN) to minimize routing errors. Here, beacons are transmitted by unmanned aerial vehicles (UA Vs) acting as ferries to inform about their succeeding anchor position to other nodes. The node receiving the beacon determines the closest ferry node and base station to the destination. Using the store-carry and forward approach, each node in the network handles sparse networks. Here, we focused to achieve low latency and have developed a novel communication protocol with two options to emphasize the need for lower latency and routing overhead when delivering data. Data collected by searching UAVs is sent to the base station using the ferry. The proposed model is validated using simulated scenarios, which show improved packet delivery, decreased end-to-end delay, and decreased overhead. The study highlights the importance of using communication methods that have lower latency and routing overhead to deliver data efficiently.
Recent Developments and Challenges in Strained Junctionless MOSFETs: A Review Amrita Kumari, Ashish Saini, Amit Kumar, Vivek Kumar, Mukesh Kumar Proceedings of International Conference on Computational Intelligence and Sustainable Engineering Solution Cises 2023, 2023 Junctionless (JL) devices represent a new generation technological development. Several studies have been conducted with an emphasis on various JL device topologies with and without incorporating strain. In this paper, an attempt has been done to provide a succinct overview of strained-JL (S-JL) devices. Moreover, a review of modelling techniques has also been provided. The need for JL architecture has also been explored for unstrained devices. To examine the improvement in device performance, simulations have been performed by several research groups. According to the study, ultra-short channel devices are in need of this technology since additional scaling to increase transistor density and device functionality has already approached its maximum limit. The ultra-abrupt junctions need not to be constructed in JL devices. A study on the benefits of strain in JL MOSFETs has also been performed. In such strained devices, improvements in ON current of up to 30-40% have been characterized. Improvements in leakage currents and threshold voltage were also seen in strained devices.
Exploratory Testing of Software Product Linesusing Distance Metrics Ashish Saini, Raj Kumar, Amrita Kumari, Satendra Kumar, Mukesh Kumar Proceedings of International Conference on Computational Intelligence and Sustainable Engineering Solution Cises 2023, 2023 Quality is very important in any software product, so that the customer trusts the efficiency of the product. The quality or validity of a product is measured by its testing. The test measures the quality of a product as well as shows the validity of the product. For example, testing and measuring the validity of a product in a single software system is not easy. Similarly, the software product line, which produces many different products of the same family at once, makes it even more difficult to check their validity and quality. Several techniques and models have been proposed to test the validity of the software line products and to ascertain the quality. In the same direction, we also propose a method for testing the products of software product line. The proposed method isbased on the distance between the features of the product (Desired and Original). We have used the proposed technique on product line with less features and found that the method presented gives fast and better results than traditional methods.
Edge Computing based IDS Detecting Threats using Machine Learning and PyCaret Amit Kumar, Sanjeev Kumar, Vivek Kumar, Amrita Kumari, Ashish Saini, Sarthak Gupta Proceedings of International Conference on Computational Intelligence and Sustainable Engineering Solution Cises 2023, 2023 The Internet of Things is getting connected with the edge computing network day by day. These devices generate large amounts of confidential data that is stored and processed on edge computing. There is a need to protect such data and edge computing networks from new types of attacks. Therefore, intrusion detection system (IDS) is an effective defense to secure edge computing network. In this study, an IDS based on machine learning is designed and developed. This article elaborates the technique of Recursive Feature Elimination (RFE) using the Random Forest algorithm (RF). The contribution of this paper is to select a relevant and efficient feature set using the recently developed Python library PyCaret. This paper proposes a novel approach for detecting new types of attacks in network traffic. The recently released dataset CICIDS2017 has been used to evaluate the proposed method, which includes state-of-the-art attacks. In this paper, two classification techniques are also used to build the IDS model. The first is a binary classification technique using the recently developed PyCaret library, and the second is a multiclass classification technique using RFE.
Strained Junctionless MOSFETs: A Brief Review Amrita Kumari, Ashish Saini, Aditya Lama, Amit Kumar 2022 International Conference on 4th Industrial Revolution Based Technology and Practices Icfirtp 2022, 2022
Classification of Minority Attacks using Machine Learning Amit Kumar, Vivek Kumar, Ashish Saini, Amrita Kumari, Vipin Kumar 2022 International Conference on 4th Industrial Revolution Based Technology and Practices Icfirtp 2022, 2022
A Review Paper on IDS in Edge Computing or EoT Reshoo Devi, Amit Kumar, Vivek Kumar, Ashish Saini, Amrita Kumari, Vipin Kumar 2022 International Conference on 4th Industrial Revolution Based Technology and Practices Icfirtp 2022, 2022
