International Journal of Innovations in Science & Technology
Not a member yet
813 research outputs found
Sort by
Comprehensive Multi-Criteria Evaluation for Landfill Site Selection in Faisalabad, Pakistan
Get PDFIntroduction/Importance of Study: Solid waste management (SWM) has become a critical issue in urban planning due to population growth and urban migration, particularly in developing countries. In Pakistan, there are no standardized regulations for landfill site selection. Faisalabad, often referred to as the "Manchester of Pakistan" due to its industrial base and growing population, faces significant challenges in this regard. Identifying a suitable landfill site is essential to minimize health and environmental risks and ensure the long-term sustainability of both urban and peri-urban areas.
Novelty Statement: This study aims to propose an optimized landfill site in Faisalabad, combining the Analytical Hierarchy Process (AHP), Multi-Criteria Decision Analysis (MCDA), and Geographic Information Systems (GIS) to guide sustainable solid waste management practices.
Material and Method: The study utilized raster, vector, and attribute data based on eight key criteria: proximity to settlements, groundwater depth, roads, airport, surface water, power stations, railway infrastructure, and population density. Using AHP within the MCDA framework and GIS modeling with weighted overlay operations, we identified potential landfill sites for Faisalabad. Population data was incorporated to validate site suitability.
Results and Discussion: Through geospatial analysis, we identified and prioritized three potential landfill sites. After a population analysis, we recommended Site-1, covering 147 acres, as the most sustainable option for the next 50 years. This site offers a balance between accessibility and environmental safety.
Concluding Remarks: The integration of AHP and GIS under MCDA proved to be an effective method for landfill site selection. These tools can significantly aid decision-makers in achieving environmentally sustainable outcomes. Future research incorporating real-time data and community feedback could enhance site selection and decision-making processes
Signal Processing Approach to Detect Non-Metallic Targets for Through Wall Imaging Using UWB Antenna
Get PDFThis paper presents a signal-processing approach aimed at detecting non-metallic targets through walls using an Ultra-Wideband (UWB) antenna configuration. Through-wall imaging holds significant importance in various fields including security, surveillance,and search and rescue operations. The proposed methodology involves the utilization of the UWB Antipodal Vivaldi antenna, known for its wide bandwidth and high-resolution imaging capabilities. The study focuses on the detection of non-metallic targets using signal processing techniques. A through-wall simulation model utilizing a UWB Antipodal Vivaldi Antenna has been developed in CST Microwave Studio. This model generates received signals in the time domain for through-wall target detection, which are subsequently processed using signal processing techniques to produce 2D images
Mininet-IDS: A Step Towards Reproducible Research for Machine Learning Based Intrusion Detection Systems
Get PDFSoftware-Defined Networking (SDN) has revolutionized network management by enabling more flexible, programmable, and controlled networks. However, the SDN controller can be a target for attacks that could bring down the entire network. In this context, intrusion detection systems (IDS) are essential for maintaining network security. Modern IDS are often enhanced with machine learning models to detect a range of network attacks. This process typically includes dataset preprocessing, model training, and integration of these models into network emulators like Mininet. However, this workflow can be complex and error-prone. To address these challenges, we present Mininet-IDS, a comprehensive command-line interface (CLI) tool that streamlines the process by offering integrated functionalities for dataset preprocessing, feature selection, model training, and deployment within the Mininet environment. Our tool simplifies the workflow by eliminating compatibility issues and ensuring reproducibility. We evaluate Mininet-IDS using the NSL-KDD dataset, training various machine learning models to detect DDoS attacks. Our results demonstrate the tool\u27s efficiency and accuracy, making it a valuable resource for network security researchers to conduct experiments with minimal machine learning expertise
Medical Intent Classification Using Ensemble and Deep Learning Models
Get PDFIntroduction: Medical chatbots are innovative solutions that leverage Natural Language Processing (NLP) and Artificial Intelligence (AI) to enhance communication efficiency between healthcare providers and patients. In the realm of conversational AI, intent classification—the task of understanding a user\u27s intent from natural language input—is both a complex and crucial aspect of the technology. This process is vital for ensuring that chatbots can accurately interpret and respond to patient queries in a meaningful and contextually appropriate manner.
Novelty Statement: This research proposes a hybrid approach that combines transformer-based embeddings with traditional deep learning models to reduce both complexity and computational cost in medical intent classification. By integrating the strengths of advanced transformer techniques with more established models, this approach aims to improve efficiency without sacrificing performance, making it more suitable for real-world healthcare applications.
Material and Method: This study investigates the use of context-aware word embeddings, including word2vec and sentence transformers, to capture rich semantic information from medical text. To refine the unstructured data, we apply various NLP preprocessing techniques, such as text cleaning, stop word removal, and lemmatization. For classification, we utilize a combination of ensemble-based and deep learning methods, including XGBoost, Random Forest, LSTM, and Bi-LSTM. These methods are tested on real-world data from 6,662 patients, with the dataset containing 25 distinct classes.
Result and Discussion: Empirical analysis demonstrates that the Bi-LSTM model, when combined with sentence transformers, achieves an accuracy of 95.23%, outperforming state-of-the-art models reported in the relevant literature.
Concluding Remarks: This research is expected to be highly beneficial to healthcare professionals by enhancing information extraction and enabling more effective handling of patient queries
Leveraging Generative AI to Learn Impact of Climate Change on Buildings Urban Areas
Get PDFClimate change, global warming, and pollution are intensifying daily. As urbanization increases, understanding the reciprocal impact between buildings and the environment becomes increasingly important. While most research on building monitoring through the Internet of Things (IoT) emphasizes energy consumption and data collection, it often overlooks the effects of outdoor environmental factors on buildings and vice versa. Additionally, existing studies frequently lack detailed reports that clarify their findings. This work aims to expand our understanding of environmental influences on buildings, indoor environments, and residents. It also seeks to generate comprehensive reports on these impacts, providing actionable recommendations to mitigate and minimize them through the use of Generative Artificial Intelligence. Specifically, we fine-tuned Large Language Models (LLMs) such as Generative Pre-trained Transformer 2 (GPT-2) and Large Language Model Meta AI 2 (LLAMA2-7b), using the Nous Research LLAMA2-7b-hf version from Hugging Face, on a custom dataset compiled from diverse online sources. Our research examines the effects of environmental factors, including temperature, humidity, and air quality, on urban buildings and indoor environments, with these models generating reports that offer practical recommendations. The generated reports offer a clear understanding of environmental impacts on buildings and suggest strategies to minimize these effects. These insights are intended to support effective urban planning and sustainable development. By implementing these recommendations or best practices, we can enhance indoor environmental quality while reducing contributions to global warming. Future work will involve continuous monitoring of buildings\u27 indoor environments, energy consumption, and greenhouse gas (GHG) emissions, further reducing GHG emissions and addressing global warming
Integrating LLM for Cotton Soil Analysis in Smart Agriculture System
Get PDFCotton is a critical crop for the agricultural economy, with its productivity closely tied to soil quality, particularly soil nutrient levels and pH. Monitoring and optimizing these properties is essential for sustainable cotton cultivation. This study proposes using fine-tuned Large Language Models (LLMs)—specifically GPT-2 and LLaMA-2—to automate soil analysis and produce detailed soil reports with actionable recommendations, addressing the limitations of traditional machine learning models in this context. A custom dataset was created by extracting key information from cotton-specific resources, focusing on soil nutrient interpretation and recommendations across different growth stages. Fine-tuning was applied to GPT-2 and LLaMA-2 models (specifically, the Nous Research version LLaMA2-7b-hf from Hugging Face), enabling them to generate data-driven reports on cotton soil health. The fine-tuned GPT-2 model achieved a training loss of 0.093 and an evaluation loss of 0.086, outperforming LLaMA-2, which had a training loss of 0.033 and an evaluation loss of 0.25. Evaluation with BERT Score showed that GPT-2 scored average Precision, Recall, and F1 scores of 0.9284, 0.9308, and 0.9296, respectively, highlighting its superior report accuracy and contextual relevance compared to LLaMA-2. The generated reports included soil properties and actionable nutrient management recommendations, effectively supporting optimized cotton growth. Implementing fine-tuned LLMs for soil report generation enhances nutrient management practices, contributing to higher yields and more sustainable cotton farming
Volatility Prediction in Cryptocurrency Using NFTs
Get PDFThe cryptocurrency market has evolved in unprecedented ways over the past decade. However, due to the high price volatility associated with cryptocurrencies, predicting their prices remains an attractive research topic. While many researchers have focused on predicting cryptocurrency prices, there has been relatively little attention given to the latest trend in blockchain applications, specifically non-fungible tokens (NFTs). In this study, we have prepared a dataset comprising NFT sales and transaction data, along with information from other cryptocurrencies. This dataset is utilized to forecast the future price of Bitcoin using several machines learning models, including Linear Regression, Random Forest, and XG Boost. The results highlight the prediction accuracy of these models. Among the three, the Random Forest regressor demonstrates the highest accuracy, followed closely by the XG Boost regressor and Linear Regression. These findings may assist investors in making informed decisions when investing in cryptocurrencies
Human Factors and Risk Analysis in Conventional System of Marble Mining: Using HFACS Framework and Structural Equation Modeling Technique
Get PDFAccidents in mines can occur due to sources of hazards that lead to the loss of hundreds of precious lives every year. Among these sources, human error is considered one of the significant sources that contribute to human errors causing accidents. In this study, different risk factors were analyzed that contribute to human errors and subsequent accidents in the conventional marble mining system. Data was collected from marble mines workers through a questionnaire based on the Human Factors and Classification System framework. Structural equation modeling was applied to examine the interaction between contributory factors that trace back to mine accidents. Two structural models were developed, showing good fit for indices with chi-square to the degree of freedom values of 2.967 and 2.095, respectively, and root mean square error of approximation value below 0.08. The results indicate that the risks caused by individuals or systems have considerable effects on human performance and safety. The findings further explore that safety management at the managerial and supervisory levels is mostly associated with systematic risks, influencing safety policies, procedures, and oversight mechanisms. However, risk caused by lack of PPE, improper machinery, and lack of training has a direct effect on workers, leading to unsafe activities. These risk factors significantly contribute to the development of unsafe conditions that increase the probability and potential severity of accidents. For improving unsafe conditions, the implementation of mechanization can effectively decrease reliance on workers, thereby minimizing human errors and ultimately enhancing safety. The findings of this study will be helpful for the assessment the surface mines safety in a better way
An Artificial Intelligence Vision Transformer Model for Classification of Bacterial Colony
Get PDFThe application of AI and machine learning, particularly the vision transformer method, in bacterial detection presents a promising solution to overcome limitations of traditional methods, offering faster and more accurate detection of disease-causing bacteria like E. coli and salmonella in water, crucial for human survival, with ongoing research to further assess its effectiveness in microbiology. This research introduces a revolutionary positional self-attention transformer model for the classification of bacterial colonies. Leveraging the proven success of transformer architectures in various domains, we enhanced the model\u27s performance by integrating a positional self-attention mechanism. We presented a novel approach for bacterial colony classification utilizing a positional self-attention transformer model. This allows the model to effectively capture spatial relationships and patterns within bacterial colonies, contributing to highly accurate classification results. We trained the model on a substantial dataset of bacterial images, which ensures its robustness and generalization to diverse colony types. The proposed model adeptly captured the spatial relationships and sequential patterns inherent in bacterial colony images, allowing for more accurate and robust classification. The proposed model demonstrated remarkable performance, achieving an accuracy of 98.50% in the classification of bacterial colonies. This novel approach surpasses traditional methods by effectively capturing intricate spatial relationships within microbial structures, offering unprecedented accuracy in discerning subtle morphological variations. The model\u27s adaptability to diverse colony shapes and arrangements marks a significant advancement, promising to redefine the landscape of bacterial colony classification through the lens of state-of-the-art deep learning techniques. The high classification accuracy attained by the model, suggests its potential for practical applications in the early diagnosis of infectious diseases and the development of targeted treatments. The findings of this study underscore the effectiveness of incorporating positional self-attention in transformer models for image-based classification tasks, particularly in the domain of bacterial colony analysis
A Comparative Analysis of BER Performance for NOMA in the Presence of Rayleigh Fading and Impulse Noise
Get PDFImportance of Study: This research investigates the integration of wired and wireless communication in Smart Grid (SG) systems, addressing the challenges posed by impulse noise and the increasing demand for bandwidth.
Novelty statement: The study explores the impact of impulse noise models on Non-Orthogonal Multiple Access (NOMA) performance within fading environments, offering insights into optimizing bandwidth utilization in multi-user SG communication.
Material and Method: Numerical simulations validate the derived closed form of the bit error rate (BER) equation, utilizing a NOMA downlink system. The performance parameters for assessing the effects of impulse noise in a Rayleigh fading channel include instantaneous signal-to-noise ratio (SNR), bit error rate, disturbance ratio, and the trade-off between spectral efficiency and energy efficiency.
Result and Discussion: The research reveals that NOMA demonstrates promising performance in SG communication despite the presence of impulse noise, with BER decreasing rapidly with increasing signal-to-noise ratio (SNR). The study highlights a performance trade-off between impulse noise and fading, emphasizing the importance of accurate SNR levels for power allocation in NOMA systems.
Concluding Remarks: This study contributes novel insights into the robustness of NOMA under realistic SG conditions, offering valuable implications for enhancing reliability and efficiency in SG communication infrastructure