International Journal of Innovations in Science & Technology
Not a member yet
813 research outputs found
Sort by
Geo-Spatial Analysis to Access Land Slide Susceptibility in Tehsil Balakot, District Mansehra, Pakistan
Get PDFLand slides are one of the recurrent natural problems that are widespread throughout the world, especially in mountainous areas, which cause significant injuries and loss of human lives, damage to properties and infrastructures. The term “landslide” is the movement of a mass of rock, debris, or earth down a slope under the influence of gravity. Landslide hazard mapping is a fundamental tool for disaster management activities in fragile mountainous terrains. The main purpose of this study was to find out landslide hazard assessment by bivariate statistical modelling and prepare an optimized mitigation map of the tehsil Balakot. The modelling was performed using a geographical information system (GIS) to derive a landslide hazard map of the tehsil Balakot. To achieve the objectives of the study, two types of variables, that is, dependent variables and independent variables, were used. The dependent variable that was selected for study was landslide occurrences. As a mandatory part of the study, the sites of previous landslides were collected from Google Earth Pro software, and a consecutive field visit was also conducted to validate the landslide sites on the ground. The Independent variables were the landslide causal factors. The causal factors that were used to achieve the objectives are Slope, Aspect, Curvature, NDWI, NDVI, Geological map, Elevation, and River distance network. The DEM data, Sentinel-2 data, and regional geologic map were used to process the landslide causal factors. The information value model was used for assessing the landslide susceptibility. The landslide susceptibility map was evaluated using the ROC curve. The result of the AUC curve was 78.71% which indicated good accuracy in the identification of the landslide susceptibility zone in a regio
Impact of Empty Nest Syndrome on Parental Mental Health: Moderating Role of Coping Styles
Get PDFWhen the kids depart, parents may experience empty nest syndrome (ENS), which is a depressing and negative emotional disturbance, and it in turn affects their mental health. According to studies, there is a need to study single parents and elderly parents who are living in shelter homes. According to attachment theory, the mental health (MH) of parents is greatly impacted by their children as an outcome of the bond between parents and their children. Coping styles assume a pivotal part in how the elderly adjust to the difficulties of (ENS) and keep up with their mental health. Thus, in the recent study moderating role of coping styles was studied. The research design that was used was a cross-sectional survey. A sample of 200 parents was collected, including single parents as well, through purposive sampling techniques. Individuals aged 60 years were included in the study. The Empty Nest Syndrome Questionnaire-Indian Form (ENS-IF), Mental Health Inventory (MHI-5), and Simplified Coping Styles Questionnaire, alongside the demographic data sheet and consent form, were administered. Collected information was analysed through SPSS and Process Macro using correlation, Regression, t-test, and moderation analysis. Future researchers can develop interventions to improve coping styles so that the mental health of empty-nest parents can be enhanced
A Robust Deep Learning Model for Early Glaucoma Detection Using Retinal Imaging
Get PDFThe Glaucoma Detection System is developed in such a way that it can enable early diagnosis of glaucoma by incorporating the latest technology with the patient-centric healthcare paradigm. It uses a user-friendly interface written in the Tkinter language and a Convolutional Neural Network (CNN) model, and is mostly useful in processing medical images. The purpose of the methodology is to democratize ocular care, focus on the insidious nature of glaucoma, and emphasize the need to have a highly accurate CNN model to detect the disease at the earliest stage. The key features are preset structures and real-time image processing, which will speed up detection and allow healthcare professionals to prioritize severe cases. The system encourages the development of multimodal integration and feedback of data in order to promote efficacy, proactive eye health, as well as the principles of fair access to care
An Enhanced Similarity Measure–Driven K-Nearest Neighbor Framework for Categorical Data Classification
Get PDFMachine learning provides effective answers to real-world classification issues by combining supervised approaches (e.g., regression, SVMs, decision trees, neural networks) and unsupervised techniques (e.g., clustering, PCA). Comparing categorical data to numerical data reveals that the former is still understudied. This study compares three variations of the K-Nearest Neighbors (KNN) algorithm, Dice Coefficient KNN (DKNN), Overlap Coefficient KNN (OKNN), and Simple Match Coefficient KNN (SMKNN) on three categorical datasets: Malware Detection, Hospital Readmission (Kaggle) and Mushroom (UCI Repository). Each variation improves classification performance by incorporating a unique similarity metric. Recall, accuracy, precision, and F1-score were used to evaluate the models. According to experimental results, SMKNN consistently performed better than the other variations, obtaining an average F1-score of 93.3%, accuracy of 88.29%, precision of 89.33%, and recall of 98%. With an F1-score of 91% and an average accuracy of 83.89%, OKNN came in second, while DKNN did worse with an accuracy of 73.74%. These results demonstrate the stability and promise of SMKNN as a dependable model for categorical data classification, highlighting its exceptional and flexible performance across a variety of datasets. The study gives useful information for identifying the best KNN variations for data-driven applications
Transformers as the Foundation of Large Language Models: A Comprehensive Review
Get PDFThe transformation of Transformer architecture has led the way into a new era for NLP, as it broke the traditional RNNs, LSTMs, Seq2Seq models, etc. As their main feature, the Revolution of Transformers was the hybridization of self-attention and multiheaded attention, which allowed the models to learn dependencies across time spans of any length through positioning methods. This resulted in a quick and efficient process for training large-scale Language Models (LLMs) that could handle the data very well and simultaneously learn the long-term dependencies. This paper is titled "Transformers as the Foundation of Large Language Models: A Comprehensive Review", and it not only reflects but also presents a critically reviewed path taken by LLMs from BERT to GPT-4 and beyond, along with the better reasoning, arithmetic, and instruction following attributed to the scaling up of architecture. The review further indicates and discusses the current concerns regarding efficiency, bias, interpretability, and domain specialization, and warns that settling these issues might dictate the fate of T-bases improvements. The authors aim through this project to provide an exhaustive comprehension of the setting in which Transformers enabled LLMs and actively directed the development of contemporary AI research
Preliminary Medical Diagnosis Using Voice-Based Urdu Language Interface
Get PDFExpert knowledge is stored in the knowledge base through an externalizing process in the form of facts, procedures, heuristics, and rules. The knowledge base helps to refine the present knowledge and insert new knowledge without recompiling a program. Medical diagnosis is one of the first knowledge-based areas in which expert system principles are applied. Almost all knowledge-based medical diagnostic systems take input symptoms in the form of text and rely on the English language. This is a hindrance to illiterate and non-native English speakers of developing countries to utilize the system, and unfortunately, Pakistan is one of them. In this connection, this paper proposed an indexing method for integrating the medical diagnostic knowledge base with a Pakistani National Language-based voice-oriented user interface for accommodating the illiterate
Delving into the Practices Involved in the Creation and Dissemination of Misinformation
Get PDFThis study investigates the authenticity of news with specific training features validating the same with specific machine-learning techniques. The contents of fake news are created to make credible information that would create mass opinions and provide a strong basis to convince the readers or confuse them utterly. The fake information is usually disseminated using numerous automated algorithms. Therefore, it is very quintessential to identify the sources and authenticity of such information. With recent advancements in information communication technology, there exists a cluster of deep knowledge from which a user intends to retrieve relevant information such as news articles. For data mining and classification tasks such as fake news classification, the approach of machine learning can be employed for effective experimentation. To address the raised issues in this study, a comprehensive and diversified dataset was required that must contain relevant knowledge with sentiment tags such as authentic and fake news. To fulfill the same, a corpus comprising over 44k authentic and fake news items is collected. Moreover, this study emphasizes news classification as fake or authentic using data mining and analytics
Early Detection and Classification of Lung Cancer using Segment Anything Model 2 and Dense Net
Get PDFLung cancer is one of the most perilous diseases worldwide with high incidence and low survival rates due to late diagnosis. Accurate detection and diagnosis of lung nodules is important for early-stage detection. Machine learning and deep learning techniques have greatly improved the precision of lung nodule segmentation and classification in Computed Tomography (CT) images. The study presents a novel approach to segmenting and classifying nodules by combining foundational models with deep learning architectures. We have used the Segment Anything Model (SAM2) to segment lung nodules and Dense Net to classify them as benign and malignant. SAM2 has been tested on the datasets using different prompts to achieve better results. Foundational Models and Deep Learning architecture’s integration significantly improved Computer-Aided Detection (CADe) and Computer-Aided Diagnosis (CADx) in medical images. Experimental results proved the effectiveness of the proposed model for early-stage detection and classification of lung nodules from CT scans. SAM2 model achieves a Dice Similarity Coefficient (DSC) of 97.87% and an Intersection over Union (IoU) of 95.82% for segmentation, and the Dense Net model\u27s classification accuracy is 97.34%. The experimental results demonstrate the performance of our model compared to existing techniques
Catalytic Performance of Electro-Oxidative Natural Manganese Sand for Ammonium Nitrogen Removal
Get PDFThe environmental risks associated with ammonium nitrogen (NH₄⁺-N) pollution have led to a growing focus on prevention. Electrochemical advanced oxidation is an effective and eco-friendly method that only requires electricity and electrolytes to remove NH₄⁺-N from wastewater. This study assesses the effectiveness of electro-oxidative natural manganese sand (NMS) in removing ammonium nitrogen under different conditions. Due to NMS’s high redox potential, it significantly enhanced the electrochemical oxidation process, increasing NH₄⁺-N removal and generating reactive chlorine species (ClO⁻/HClO) when NaCl was added. The experiment was also conducted without a catalyst, quartz sand, and natural manganese sand, but NMS removed 86.4% of NH₄⁺-N, outperforming the other treatments. The removal efficiency was tested at five different pH levels (3, 5, 7, 9, and 11), with NMS showing the highest efficiency of 95.2% at pH 9. At a current density of 15.5 mA/cm², the removal rate reached 94.9%, and with a NaCl concentration of 9 g/L, the removal efficiency peaked at 96.2%, driven by increased production of reactive chlorine species (ClO⁻). These results demonstrate the electro-oxidative NMS system as a highly efficient, scalable, and eco-friendly solution for ammonium nitrogen removal in wastewater treatment
Improving Software Requirements Elicitation in Agile Environment
Get PDFRequirement elicitation plays an important role during the software development life cycle. The selection of an improper requirement elicitation method will affect the quality of developed software. Agile methodologies are popular in the industry and follow an incremental approach to developing software. Agile methodologies value customer needs, interaction among teams, interaction with customers, and change management. Researchers proposed methods for requirement elicitation in agile software development. This research aims to investigate the issues faced during requirement elicitation in agile software development. We will identify the method that motivates the requirements elicitation in agile software development to meet our objective. After identifying the literature, a systematic literature review will be performed. An introductory overview, publications trends and values, strengths, and limitations will be highlighted. Based on the identified limitations, we will propose a new requirement elicitation method useful in agile software development. To evaluate the results, two teams of equal expertise were given the same project to develop. One of the teams developed using the proposed framework, and the other one did without using the proposed framework. Then, both of them were given the survey they filled out and gave their input on the requirement elicitation parameters, and the results were compared and validated using a t-test and reliability analysis