Ajay Kumar Singh

Ajay Kumar Singh

Auto-generated profile, based on publicly available information

Ajay Kumar Singh

Formerly Uttar Pradesh Technical University

Research Focus & Metrics

Academic Career Length

14Years (2013-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 February 2023 in Indonesian Journal of Electrical Engineering and Computer Science
Verified
Journal Article
Open access
PDF file
NBnidhi bansal
nidhi bansal
ASAjay Kumar Singh
Ajay Kumar Singh
Makespan and cost are the major factors in the current scenario and they generally upgrade the results to optimize the upcoming task processing by implementing the scheduling within the specified cloud platform. A new proposed agenda is being considered for a health care system to make the world healthier. The paper is designed to identify prior work as a health evaluator for the end user and act accordingly. It also satisfies the end user by providing demanded results and can establish trust between the system and the customer within a global era. By analyzing the COVID-pandemic situation in the digital world, things should be implemented in conjunction with healthcare and technology to serve things better. The proposed algorithm works on priority basis by analyzing the current patient condition and then implemented in CloudSim Toolkit. As per the results, the proposed steps are performing 50-70% better in terms of makespan and cost. Notable optimization has been accomplished by the proposed healthcare evaluator.

1 January 2022 in Indonesian Journal of Electrical Engineering and Computer Science
Verified
Journal Article
Open access
PDF file
NBnidhi bansal
nidhi bansal
ASAjay Kumar Singh
Ajay Kumar Singh
Quality-based services are an indicative factor in providing a meaningful measure. These measures allow labeling for upcoming targets with a numerical comparison with a valid mathematical proof of reading and publications. By obtaining valid designs, organizations put this measure into the flow of technology development operations to generate higher profits. Since the conditions were met from the inception of cloud computing technology, the market was captured heavily by providing support through cloud computing. With the increase in the use of cloud computing, the complexity of data has also increased greatly. Applying natural theory to cloud technology makes it a fruit cream. Natural methods often come true, because survival depends on the live events and happenings, so using it in real life as well as any communication within technology will always be reliable. The numerical results are also showing a better value by comparing the optimization method. Finally, the paper introduces an adaptation theory with effective cloudsim coding of honey bees and grey wolf in conjunction with their natural life cycle for solving task scheduling problems. Using adapted bees improved the results by 50% compared with the original bees and secondly by honeybees and grey wolf improved 60%.

3 April 2013 in International Journal of Image, Graphics and Signal Processing
Verified
Journal Article
Open access
PDF file
DCDolly Choudhary
Dolly Choudhary
ASAjay Kumar Singh
Ajay Kumar Singh
STShamik Tiwari
Shamik Tiwari

26 May 2026 in Advanced International Journal for Research
Verified
Journal Article
Closed access
ASAjay Singh
Ajay Singh
KBKalyanam Bharat
Kalyanam Bharat
SBShyam Baldua
Shyam Baldua
HJHarshil Jain
Harshil Jain
This paper introduces a conversational medical assistance system that integrates a Large Language Model with retrieval-based grounding mechanisms to provide structured, non-diagnostic health guidance. Unlike conventional symptom checkers that rely on rigid decision trees, the proposed system dynamically interprets user queries, retrieves evidence-backed medical information from curated sources, and synthesizes contextual recommendations. The assistant generates medication suggestions, dosage information, safety precautions, lifestyle advice, and nearby pharmacy references. By incorporating Retrieval-Augmented Generation (RAG), the architecture reduces hallucinated responses and enhances factual alignment. Experimental validation across common minor conditions demonstrates improved contextual coherence and structured output reliability. The system is designed with a modular architecture that separates natural language processing, knowledge retrieval, and response generation layers to ensure scalability and maintainability. The retrieval module leverages indexed medical datasets and trusted clinical references to fetch relevant information in real time, while the language model refines this information into clear, user-friendly explanations. The interface supports conversational interaction, enabling users to ask follow-up questions and receive context-aware responses without repeating prior information. Furthermore, safety constraints are embedded within the system to prevent diagnostic claims and to encourage users to seek professional medical consultation when necessary. Performance evaluation metrics such as response accuracy, contextual relevance, and information completeness indicate that the integrated approach provides more reliable and interpretable outputs compared to standalone generative systems. This framework highlights the potential of combining modern AI language models with validated knowledge sources to develop responsible and practical digital health assistance tools. The proposed solution emphasizes interpretability and transparency as key design priorities, ensuring that users can easily understand the reasoning behind generated responses. By structuring outputs into clearly defined informational segments, the system minimizes ambiguity and promotes clarity in health-related communication. The conversational interface is also designed to accommodate natural language variations, allowing individuals to describe symptoms using everyday expressions without requiring technical medical vocabulary. This capability enhances accessibility for users from diverse backgrounds and improves overall usability of the platform. Moreover, the integration of retrieval-grounded generation establishes a balance between linguistic flexibility and factual reliability, which is essential in domains where informational accuracy is critical. The architecture demonstrates how modern AI systems can be adapted to provide supportive guidance while respecting safety boundaries and ethical considerations. As advancements in medical datasets, language modeling techniques, and retrieval optimization continue, such hybrid systems are expected to evolve into increasingly dependable digital tools capable of assisting users in making informed health-related decisions while complementing, rather than replacing, professional medical services.

29 May 2026 in Journal of Internet Services and Information Security
Verified
Journal Article
Closed access
JRJeevaraj R
Jeevaraj R
ASAjay Kumar Singh
Ajay Kumar Singh
As counterfeit drugs continue to become more prevalent in the pharmaceutical supply chain, it poses significant threats to patient safety, public health, and economic stability. The traditional pharmaceutical supply chain systems are primarily based on centralised databases and manual checks, which limits transparency, traceability and data integrity. According to various global health organizations, counterfeit drugs make up more than 10% of the pharmaceutical products in developing nations, and great need of a secure and transparent monitoring system. The goal of this study is to create a framework for a blockchain and Internet of Things (IoT) enabled pharmaceutical supply chain that can enhance the tracing and security of pharmaceuticals. It combines the technology of distributed ledger with IoT sensing devices to create a secure transaction validation, real-time monitoring and tamper-resistant tracking system in the supply chain. A mathematical model that is based on this probabilistic aspect of the sensing reliability and verification frequency is developed for the evaluation of the probability of detecting a counterfeit drug. Moreover, an adversarial behavior and economic incentives analysis using a game-theoretic security model is provided to help understand the CTD problem. The real-time logistics monitoring and secure cloud verification are implemented using edge devices based on AM62x. Experimental results indicate that the proposed framework can achieve a higher probability of counterfeit detection (up to 0.89-0.98) compared with the conventional systems (0.48), and can reduce the transaction confirmation latency from 250 ms to 80 ms. The results indicate that there is greater transparency, trust and resiliency in the pharmaceutical supply chains.

19 March 2026 in Advanced International Journal for Research
Verified
Journal Article
Closed access
RRRiya Raju
Riya Raju
ASAjay Singh
Ajay Singh
The proliferation of fake news through online channels is a real menace that has posed considerable problems in determining the authenticity of online information. This is a multimodal misinformation de-tection project that is an integrated project using both text and visual data in order to find out misleading information in a more effective way. The system uses transformer-based models including BERT, RoB-ERTa and XLNet to perform text classification besides CNN and FastText models to receive semantic and contextual features of the news articles. To verify images, convolutional neural networks and mobile net are used in order to identify manipulated or tampered pictures. This compilation of models makes it possible to create a single framework that will increase the accuracy of detection in a variety of data sets. Besides, Explainable AI methods, in particular SHAP, are included to explain decisions of a model and visualize the importance of features to guarantee transparency and trust in predictions. The entire system is deployed as a web application in Flask, whereby it has user registration modules, log in modules, con-tent submission modules, and modules to display the classification results. The findings of the experi-ment prove that the suggested scheme has a better reliability level than single-modality systems. On the whole, the current project allows creating a viable and interpretable misinformation detection system based on natural language processing, computer vision, and explainable AI.