LC International Journal of STEM (ISSN: 2708-7123)
Not a member yet
120 research outputs found
Sort by
Plant Leaf Disease Detection Using Deep Learning
Get PDFPlant leaf diseases pose a danger to food security, and their rapid identification is made more difficult in many areas by a lack of infrastructure. This thesis is a concentrated attempt to address this important problem by utilizing state-of-the-art deep learning techniques, with a focus on the YOLOv5 model, to offer a dependable and effective solution for plant leaf disease detection in agriculture. The introduction emphasizes the serious effects that plant diseases have on a global and financial level, underscoring the critical necessity for early detection to lessen these effects. Driven by the promise of technology to revolutionize agriculture, this work carefully investigates the complex use of deep learning techniques. YOLOv5 is trained to demonstrate its ability to distinguish between healthy and diseased plant leaves using a carefully chosen tomato dataset. The dataset contains nine different types of illnesses. The model achieves an impressive 92.6 percent average precision, indicating a high degree of disease detection accuracy. Plant leaf disease detection in agriculture faces many complicated obstacles, and the successful deployment of the trained model through the Flask framework represents a significant leap in the practical application of deep learning to address these issues. Our multimodal approach places our research at the forefront of efforts to improve agricultural technology and guarantee global food security while also making a significant contribution to the scientific understanding of disease identification and laying the foundation for future advances
Service Quality of Higher Education in the Selected Govt. Colleges of Northern Bangladesh: An Evaluation on Servqual Dimensions
Get PDFTo speed the rate of higher education government (govt.) has increased and promoted the number of govt. degree level institutions as higher level honor’s and master’s institution. There is no measure to ensure quality of these higher educational institutions as prescribed in the public universities or international universities. To know service quality and performance of the service provider groups measuring service quality is very much significant. For this purpose, we have to measure students’ expectations from the service entity and their perceptions about the actual services they get. The govt. colleges that provide higher education need to understand the customers’ (students’) perceptions of service quality and identify the gaps between their expectations and these perceptions. This part of the research studies students’ expectations and perceptions of service quality in the present educational environment by using the modified service quality (SERVQUAL) model given by Parasuraman et.al for distinctive nature of service. The modified SERVQUAL instrument will measure five dimensions i.e. tangibles (academic, non-academic), reliability, responsiveness, assurance and empathy. This study has been done over 382 undergraduate and post graduate students of selected four govt. colleges from top ten ranking 2016 by National University of northern division Rajshahi and Rongpur. Next the descriptive statistical analysis with mean scores, standard deviation and crosstabs in different perspective by colleges and respondents with service dimensions. Service quality has been measured by subtracting students’ expectations mean scores from their perceptions mean scores of different service items related to educational services based on SERVQUAL dimensions. A structured questionnaire has been supplied with same type questions to know students’ expectations and perceptions with a 5 point Likert type scale. The study found that all the SERVQUAL components has made negative scores of service quality of higher education in govt. colleges
A Comparative Analysis On Cleveland And Statlog Heart Disease Datasets Using Data Mining Techniques
Get PDFIn today’s age deaths due to cardiovascular diseases are turn out to be a major problem. Factors such as high blood pressure, diabetes, high cholesterol level, hypertension, smoking and obesity are high risk to cause cardiovascular disease. Many researchers are using different datasets of heart patients to early diagnose the cardiovascular disease such as Cleveland and Statlog heart disease dataset. This study aims to compare the results of previous studies using Cleveland and Statlog heart disease datasets. We analyzed that different machine learning and deep learning techniques had been applied on these datasets which showed different resultson Cleveland and Statlog datasets
Tourism Sector and Indian Economy: An Evaluation from 2000 Onwards
Get PDFIn this research paper it is illustrated how the travel and tourism industry boosts the Indian economy. Tourism can be defined as discovering explored or unexplored destinations. It comprises of understanding a different lifestyle, relishing the mouth-watering dishes and rejuvenating oneself. The Indian government has realized this and is therefore working towards achieving a target of 1% inbound tourist arrivals worldwide to India by 2021. The Government of India has started infusing funds into the industry through various schemes and projects like the Swadesh Darshan, Pilgrimage Rejuvenation and Spiritual Augmentation (PRASAD) etc. With the rise in stress and competitive lifestyle, people are ardently searching for alternatives to hit the road and rejuvenate themselves. They mostly opt for religious places as they feel that their, they will receive the peace and calmness they are looking for. There has been a 14% growth in the foreign tourist arrivals in India in 2017. Tourism has significantly helped the country by contributing majorly to the GDP, providing large-scale employment, promoting skilled and unskilled labor and by being a major source of foreign exchange earring. It is quite evident that the travel and tourism industry has achieved a positive growth and that it will further expand to achieve great height
Enhancing Face Recognition through Dimensionality Reduction Techniques and Diverse Classifiers
Get PDFFace recognition is essential component of various applications including computer vision, security systems and biometrics. By examining the efficacy of several dimensionality reduction techniques, including Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA), Singular Value Decomposition (SVD), and Non-Negative Matrix Factorization (NMF), this paper offers a novel approach to face recognition. These techniques are combined with diverse classifiers, including Support Vector Machines (SVM), Random Forest, LightGBM, and k-Nearest Neighbors (KNN), are employed to evaluate their impact on face recognition accuracy. Experiments were conducted on Olivetti faces data set. We have demonstrated the comparative analysis of different dimensionality reduction techniques classifiers in terms of accuracy, precision, recall, f1score. Results shows the potential of integrating PCA with diverse classification models to enhance face recognition accuracy and highlights its applicability in real-world scenarios. 
Enhanced Efficiency and Productivity through AAMS
Get PDFThe Traditional attendance management systems, which rely on human operations or RFID-based solutions, frequently struggle with scalability, accuracy, and efficiency. This thesis proposes an Automated Attendance Management System (AAMS) that employs a customized YOLOv9-C model for real-time facial recognition via deep learning. The model's performance is significantly improved by adding Squeeze-and-Excitation (SE) blocks and the Complete Intersection over Union (CIoU) loss function. On a custom dataset, the baseline YOLOv9-C model had 86.2% precision and 84.9% recall, with a mean Average Precision (mAP) of 89.9% at IoU threshold of 0.5. However, the revised YOLOv9-C(M) model demonstrated significant gains, including a mAP of 93.8%, as well as improved precision (94.1%) and recall (96.6%).
These improvements can be due to the introduction of SE blocks, which promote feature recalibration, and the CIoU loss function, which maximizes bounding box localization and increases detection accuracy even in tough conditions such as occlusion or dimly lit areas. The improved YOLOv9-C model consistently outperforms the existing YOLO models (YOLOv5, YOLOv7, and YOLOv8s), according to a comparison study. The mAP for YOLOv5 was 80.2%, YOLOv7 was 89.1%, and YOLOv8s was 91.4%. In contrast, the upgraded YOLOv9-C model outperformed the others, with greater robustness, precision, and recall.
The system employs a one kind of custom dataset to evaluate the model's performance in some scenarios and settings, as well as to ensure trustworthy workforce detection in diverse contexts. By automating the attendance process, this technology reduces errors, saves administrative time, and promotes institutional efficiency
Literature Review on X-Ray based Pneumonia Detection using Machine Learning and Deep Learning Methods
Get PDFArtificial intelligence has proven to be an effective way in the detection of many diseases. This study presents a literature review of artificial intelligence techniques used in the detection, classification and visualization of pneumonia disease in lungs using radiographs of chest. In this review, different reliable databases were searched including research gate, ELSEVIER, Applied sciences and IEEE. Pneumonia is a fatal sort of malady on the off chance that we truly couldn't care less. If we don’t diagnose it in its early stages it can be responsible for 50000 deaths every year [59]. There are two kinds of pneumonia: viral and bacterial. Many researchers have done their research for the identification of pneumonia using machine learning and deep learning methods. This study gives you an overview of the machine and deep learning methods proposed previously for the pneumonia detection. The review is structured based on Deep learning, transfer learning and machine learning methods using chest x-rays images for the early identification of pneumonia. The main objective is to find the limitations of the previous studies and suggestions for the future work
Design and Implementation of Hostel Management System Using Java and MySQL
Get PDFThe Hostel Management System framework is a software which is design to provide the facilities to the staff members as well as the students that saves the time that required by those paper works. As many students willing to live at hostel for studies the number of hostel buildings are increased that needs to be handle smartly by using the web application that decreases the stress or strain to the authorities. This application requires students and administrators login details to take them over the application dashboard where they all can easily access the information regarding their registration for hostel rooms, fee payments, can check student records as well as allow to update whenever required. This application overcomes the drawbacks of the past methods of management system; it is user friendly, GUI interface or environment, reliable and secured with best IT department professionals. Working on such application brings transparency in the environment that builds the trust between students and management as well as trust on the digital world
Detecting Mobile Money Laundering Using Genetic Algorithm as Feature Selection Method with Classification Method
Get PDFIn recent years, mobile phone payment systems have been extensively used in developed countries. Frauds are affecting the economy of the whole world. Different kinds of mobile money frauds are credit card, bank fraud, insurance fraud and financial fraud. In this paper we discussed financial fraud and proposed an effectiveness method for money laundering. Payment system in fraud divided into four parts, point of sale, mobile payment platform, mobile payment independent and bill payment through mobile. Mobile phones are great source of service for financial transactions. Our objective is to identify the misuse of mobile money transaction and to prevent fraud from financial transaction to save the money. Financial Action Task Force (FATF) is an organization that views internationally money laundering. Financial Action Task Force continuously strengthens its standards for dealing with new risks.The Financial Action Task Force monitors countries to ensure the implementation the Financial Action Task Force Standards and holds countries to account that do not comply. This paper proposes hybrid Genetic algorithm based on feature selection method and investigates the performance of Decision Tree and Boost classification machine learning method. We applied Area under the ROC curve (AUC) and confusion matrix after using the feature selection method. We found the resultsof Decision Tree validation, testing and Boost with different Sampling of both datasets and Boost has better performance than Decision Tree
Understanding Public Opinions on Social Media about ChatGPT – A deep Learning Approach for Sentiment Analysis
Get PDFUser-generated multimedia content—photos, text, videos, and audio—is becoming more and more common on social networking sites to allow individuals to express their thoughts. One of the largest and most advanced social media platform discussing ChatGPT is Twitter. This is because Twitter updates are constantly being produced and have a limited duration. The deep learning method for sentiment analysis of Twitter data about ChatGPT evaluation is presented in this research. This study used 4-class labels (sadness, joy, fear, and anger) from public Twitter data stored in the Kaggle database. The proposed deep learning strategy significantly improves the efficiency metrics determined by the use of the attention layer in current LSTM-RNN approaches, increasing accuracy by 20% and precision by 10-12%, but recall only 12-13%. Out of 18000 ChatGPT-related tweets, positive, neutral, and negative sentiments accounted for a respective 45%, 30%, and 35%. It is determined that the suggested deep learning technique for ChatGPT review sentiment categorization is effective, realistic, and fast to deploy