January 2025 Volume 11 Issue 1

ENGINEERING & TECHNOLOGY JOURNAL

Algorithm Performance Comparisons: A Study on Heart Disease Prediction Models

Shweta Tripathi, Neha Goyal, Mohammad Faisal

  • Pages: 1-8
    • <p>Heart problems rank among the world's major causes of death. Therefore, improved predicting algorithms must be created. This study analyses a variety of algorithms using machine learning (ML) to improve the accuracy of heart disease prediction. We focus on frequently used algorithms such as neural networks (NN), decision trees, random forests, logistic regression (LR), and support vector machines (SVM). The performance of these algorithms is measured using metrics such as F1-score, recall, accuracy, and precision. This study uses a dataset from the UCI Machine Learning Repository. The findings demonstrate significant variances in the algorithm's effectiveness indicators, which will inform future research and therapeutic applications.</p>

Brand Reputation Management from Social Media Sentiment Analysis using Machine Learning: Survey on the Standard Processes

Dayapa Srilatha, Dayapa Srilatha, Harikrishna Bommala

  • Pages: 1-8
    • <p>Effective brand reputation management is essential in the current digital era, since social media platforms play a pivotal role in facilitating brand-consumer interactions. This study explores the field of brand reputation management using social media sentiment analysis with the use of machine learning techniques. As social media platforms continue to grow in number, it has become more challenging yet essential for businesses to monitor and understand how users feel about their brand. This is crucial for organizations who want to uphold and improve their reputation. The article presents a complete framework that utilizes machine learning algorithms to assess attitudes posted on different social media sites. The system utilizes natural language processing (NLP) techniques to derive significant insights from the large volume of textual data produced by users. The framework's objective is to utilize sentiment analysis to identify the dominant attitudes, opinions, and emotions linked to a brand's products or services. The process entails gathering data from several social media channels, including platforms such as Twitter, Facebook, Instagram, and online forums. The system utilizes machine learning models, including Support Vector Machines (SVM), Naive B ayes, and Recurrent Neural Networks (RNNs), to analyze textual input and categorize attitudes as positive, negative, or neutral. Moreover, methods like topic modeling and entity identification are utilized to detect prominent topics and important entities in user-generated material.The study finishes by examining the ramifications of utilizing social media sentiment analysis in the context of brand reputation management. Although technology is effective in monitoring and understanding user attitudes, the human element is still crucial in providing context to insights and developing strategic solutions. The combination of machine learning algorithms and human knowledge creates a mutually beneficial strategy to effectively manage brand reputation in the digital age.</p>

An Automatic Method for Detecting Brain Tumor Tissue in T-1 weighted MRI images

Saumya Singh, Sumit Yadav, Shweta Dwivedi , Vishal Agrwal

  • Pages: 1-4
    • <p>Brain tumors are a critical health issue, particularly among individuals aged 0 to 19, where they are the predominant type of cancer. Prompt and precise diagnosis is essential as these tumors are the leading cause of cancer-related deaths in this age group. Magnetic Resonance Imaging is a crucial tool for detecting brain tumors due to its ability to provide detailed and accurate images. However, the manual analysis of MRI scans can be challenging due to the variability in tumor appearance and the complexity of brain structures. To address these challenges, advanced biomedical image processing techniques have been developed. These techniques typically involve a series of steps to enhance the accuracy and efficiency of tumor detection. A comprehensive approach consists of four key stages: feature extraction, morphological operations, segmentation, and classification. 1. Feature Extraction: This stage involves identifying and extracting relevant features from MRI scans that can help differentiate between normal and abnormal brain tissues. 2. Morphological O perations: These operations are applied to refine the images, improving the clarity of the tumor boundaries by removing artifacts and noise. 3. Segmentation: This step is crucial for isolating the tumor from the surrounding brain tissue. It involves dividing the MRI image into distinct regions to accurately locate and outline the tumor. 4. Classification: Finally, the segmented regions are classified to determine the presence and type of tumor. This step often employs machine learning algorithms, such as Support Vector Machines, to enhance the accuracy of tumor identification. To improve the segmentation process, noise reduction techniques like median filtering are used to clean the MRI images, and brain extraction methods are applied to remove non-brain structures such as the skull. These preprocessing steps are critical in ensuring that the subsequent analysis focuses solely on the brain tissue. Overall, automating the analysis of MRI scans through these advanced techniques helps in early and precise detection of brain tumors, ultimately aiding in better management and treatment of the condition.</p>

A Novel Approach to Heart Disease Diagnosis Using ML

Shesh Kumar , Sachin Kumar Sonker , Bhanu Pratap Rai , Divya Singh ,Lalit Kumar Tripathi, Ajai Kumar Maurya

  • Pages: 1-10
    • <p>an immense number of Indians still lack access to topnotch, priced affordably medical treatments. Morbidity and death increase because the disease has advanced when medical diagnostic and therapy guidance is put off during its initial phases of the malady... In India, the death rate from non-communicable illnesses has risen dramatically during the previous two decades. Deaths from cardiovascular diseases [CVDs] have skyrocketed over these two time periods, rising from 15.2% to 28.1%. During the past two decades, not only has mortality from cardiovascular disease [CVD] skyrocketed, but so has mortality from other chronic illnesses including cancer, hepatitis, diabetes, chronic renal disease, etc. The current situation in India necessitates the use of machine learning methods to increase the availability and affordability of healthcare. The primary goal of this study is to further medical progress through the application of machine learning methods. In this paper, we propose using a sliding window approach for feature selection to zero in on the most important noninvasive clinical features for cardiovascular disease prognosis. Accuracy, specificity, and sensitivity may be maximized by finding the input properties that work well together. Using these key characteristics, a Machine-learning-driven cardiac disease wagering mechanism orchestrated to meet an accuracy of 93.8%.</p>

Coffee Leaf Disease Detection Using YOLOv8m

Shivam Kumar , Nishant Kumar , Saloni Kumari , Anish Rudra and Bappaditya Das

  • Pages: 1-4
    • <p>The early detection of plant diseases is crucial for preventing crop losses and ensuring agricultural sustainability. This study explores the application of YOLOv8m (You O nly Look O nce, version 8 medium), a state-of-the-art deep learning object detection algorithm, for image classification and disease detection in coffee leaves. The Coffee Leaf Disease Dataset, which includes annotated and unannotated images of healthy and diseased coffee leaves affected by conditions like rust and miner infestations, is used to evaluate YOLOv8's performance. YOLOv8's anchor-free architecture and real-time detection capabilities make it ideal for this task, offering high validation accuracy (91.4%) and F1-confidence score (0.96) in detecting disease-specific lesions. The model demonstrates significant improvements in both speed and precision compared to traditional approaches. This research highlights the potential of YOLOv8 in agricultural disease management and underscores its value in real-time disease detection applications in precision farming.</p>
other Journals