TEJAS Journal of Technologies and Humanitarian Science

Paper 1: AI-Driven Predictive Analytics with the Help of IoT for Organizational Change Management

Authors : Esha Srivastava, Shraddha Yadav and Mahesh Kumar Tiwari

Doi: https://doi.org/10.63920/tjths.43001

Abstract

Artificial Intelligence (AI) and Internet of Things (IoT) are increasingly finding their way to the current workplaces thereby transforming Organization Change Management (OCM). The present paper discusses the opportunities of transforming the way organizations should handle change through the integration between AI-driven predictive capabilities and data measured in real-time through the IoT devices. The conventional change models are usually built on basis of stagnant planning and instinctive decisions with the AI and IoT empowering dynamic decisions and in light of data. IoT Sensors are able to record real-time behavioral, physiological, and environmental data that is processed by an AI system to analyze the data to find potential signs of resistance, disengagement, or stress. These insights can be used to intervene in a timely and personalized manner that helps in facilitating the transitional process and aiding in employee wellbeing. Based on the case studies and analyses, this paper demonstrates the opportunities and the ethical implications of such technologies in change management practice. It also ends with propositions on how to create more versatile, individualized, and morality-based change plans in future.

Paper 2: Advancement of HTTP and Security Issues Over It

Authors : Nitya Nand Dwivedi, Dharmendra Singh

Doi: https://doi.org/10.63920/tjths.43002

Abstract

Now days so many organizations are focusing that how to improve http and security related problem over it. The new protocol, HTTP / 2, has most popular in the web industry as its typical information was confirmed and accepted earlier few years. Many of its technical structures are derived from the Google application-based protocol, HTTP / 2 solves many bugs and variations of HTTP / 1.1, improving web performance during page loading times. Expected. HTTP / 2 introduces topic compression, which allows multiple simultaneous exchanges on the same connection, allowing better utilization of network resources and reducing latency. We will also introduce one-sided submissions from the server to the client. Flow and priority control can help you use more live streaming. Flow control helps ensure that only data received from the recipient is sent. Priority allows the index of limited resources to the most significant load first. HTTP / 2 adds a novel communication method that allows the server that may be send feedback to the client. This research article shows that this functionality learns the impact of HTTP / 2 performance on standard websites, and while using HTTP / 2 improves page load speed and efficiency, increasing cookie usage makes it less secure. Therefore, you need to prevent this and ensure that the cookies you use are securely stored in your system and encrypted in some way. You can allow encryption locally, but you can also re-use the website upload server. In my research I used encryption via Node.js and the XAMP server.

Paper 3: A Comprehensive Review on Tomato Leaf Disease Detection using Deep Learning Techniques

Authors : Pankaj Kumar Gupt, Dr. Anita Pal

Doi: https://doi.org/10.63920/tjths.43003

Abstract

Tomato cultivation is susceptible to various diseases, leading tocsignificant yield loss and economic impact. Rapid andcaccurate prediction is essential for timely intervention and mitigation. Deepclearning techniques, specifically CNN for the automated detection of tomato leaf diseases. The proposed methodology involves the acquisition of high-resolution images of tomato leaves, and training a CNN model to classify images into healthy or diseased categories. The dataset used for training and evaluation consists of labeled images encompassing early blight, late, along with healthy leaves. The CNN architecture is optimized through experimentation to achieve in terms of accuracy, precision, recall and F1-score. The trained model demonstrates promising results in accurately identifying various tomato leaf diseases, even in the presence of environmental variations and leaf deformities. Furthermore, the computational efficiency of the proposed approach allows for real-timecor near real-time disease detection, facilitating timely agricultural interventions. Overall, this research contributes to the advancement of automated agricultural monitoring systems, aiding farmers in early disease detection and management, thereby enhancing crop productivity and sustainability.

Paper 4: A Comprehensive Review of Machine Learning Techniques for Brain Tumour Classification and Detection

Authors : Shalini Verma, Dr. Anita Pal

Doi: https://doi.org/10.63920/tjths.43004

Abstract

Because brain tumours vary widely in size, location, and form, diagnosing them can be extremely difficult. Although manual evaluation and conventional imaging techniques are still widely used, deep learning has become a game-changing technology for automated diagnosis. The study discussed in the thesis is summarised in this review, which also places it in the larger context of brain tumour detection methods. It addresses classical machine learning algorithms, the advent of convolutional neural networks (CNNs), and hybrid procedures. The report offers a thorough reference for audiences in academia and medicine by highlighting present strengths, enduring constraints, and prospects for further research.

Paper 5: Robust Medical Image Prediction via Adaptive Reconstruction: Bridging the Gap in Low-Quality Data

Authors : Ravi Krishan Pandey, Praveen Kumar Shukla, Dharmendra Lal Gupta

Doi: https://doi.org/10.63920/tjths.43005

Abstract

Medical image prediction plays a very significant role in clinical decision-making and early detection and diagnosis of different diseases. However, the quality of medical images has a huge impact on the predictive models' accuracy. Poor-quality data usually occurs due to problems like noise, artifacts, and low resolution and poses a major challenge for reliable medical image prediction. This study develops a new framework of robust medical image prediction by exploiting adaptive reconstruction techniques that reduce the gap in low-quality data. Our method combines state-of-the-art image processing methods with machine learning algorithms to enhance the quality of medical images before feeding them into predictive models. The adaptive reconstruction-based model consists of using classic denoising techniques in images and deep learning-based approaches, selectively enhancing critical features and removing noise. It aims to provide qualities in image reconstruction suitable for prediction tasks by recovering lost or degraded information. In addition to this, the work also focuses on the use of robust machine learning algorithms to enhance prediction accuracy on the reconstructed images. The framework was tested on various datasets and had significant improvements in predictive performance when compared to the traditional approaches using low-quality images directly. The results showed that adaptive reconstruction not only boosts the visual quality of medical images but also promotes the overall predictive model performance for clinical applications. This paper provided a promising approach to overcoming such limitations from data of low quality, which will promote more accurate and reliable predictions toward clinically relevant outcomes in medical imaging.

Paper 6: Explainable AI (XAI) Techniques to Enhance Cancer Diagnosis

Authors : Harmeet Khera, Nikhil Pandey and Dr. Shalini Lamb

Doi: https://doi.org/10.63920/tjths.43006

Abstract

This study examines how Explainable Artificial Intelligence (XAI) enhances diagnosis of cancer, emphasizing the importance of trust and transparency in the clinical sector. While artificial intelligence (AI) models have demonstrated impressive accuracy in detecting cancer from medical images, their black-box nature often prevents doctors from understanding how predictions are made. This lack of interpretability creates hesitation in adopting AI for real-world healthcare. The paper examines popular XAI methods which provide visual and feature-based explanations of model outputs. It also discusses how these methods can enhance clinician confidence, reduce diagnostic errors, and meet regulatory requirements for accountability. This research emphasizes the promise of interpretable models in connecting the precision of machine learning with the trustworthiness of medical practices, based on a review of recent studies and a suggested framework for applying XAI in oncology. Additionally, it discusses future strategies for incorporating XAI into clinical processes.