Note : All Projects are Solo based projects not a Group Project
SDLC Based Projects
Object Segmentation using Google Earth
Object Segmentation in GE is the process of identifying and locating objects in an image or video using computer vision techniques. GE uses advanced algorithms and machine learning to enable object detection in various applications. The goal is to accurately and efficiently identify and track objects to improve safety, enhance productivity, and enable new capabilities in industries such as surveillance, autonomous vehicles, and robotics.
[Code] [Report]Blockchain Voting System
The Blockchain Voting System Project is a secure and innovative approach to voting that uses blockchain technology to ensure the accuracy and integrity of the voting process. By leveraging a decentralized and immutable ledger, the system eliminates the risk of voter fraud and hacking, and provides a transparent and automated voting process. The system is accessible and user-friendly, making it an ideal solution for remote and absentee voting. This system is made by using Hardhat, Ethereum smart contracts, and React would be a decentralized platform for voting. The system would use cryptography to ensure secure, anonymous, and auditable votes, with a front-end built in React and the back-end using Ethereum smart contracts, deployed with Goerli Testnet.
[Code]Machine Learning Based Projects
PMV-PPD-Prediction
The PMV and PPD prediction project uses algorithms to accurately assess thermal comfort in various settings. It predicts key factors such as temperature, humidity, and air velocity using Support Vector Regression (SVR) algorithms, and classifies thermal sensation levels using semi-supervised algorithms like KMeans clustering with logistic regression. The project is scalable and adaptable to different sensor and data collection methods, providing a responsive and nuanced assessment of thermal comfort.
[Code][Report]Medical Fraud Detection
In the United States, the median range of healthcare offenders was between 400 and 430 in 2021. While the number of offenders is relatively small, the related financial impact is significant. We utilize the Centers for Medicare & Medicaid Services (CMS) Big Data provider datasets to identify instances of fraudulent activity within the healthcare system. As of 2023, the number of datasets available has been increased, enabling more robust and comprehensive analysis. In addition, we take into consideration various treatment and Medicare Programs to ensure a thorough understanding of the healthcare landscape and to effectively identify fraudulent activity. we employ various Supervised, Unsupervised, and Semi-Supervised Algorithms to effectively not only by identify fraudulent activity within the healthcare system and also depicting where exactly did the fraud happen. Specifically, we leverage Deep Neural Networks to capture the interactions between the features within the dataset, achieving an accuracy score of 85%
[Code][Report]Leveraging ML techniques in the subject of Reverse Engineering of Malware
Reverse engineering malware is the process of decompiling and analyzing malicious software in order to understand its behavior and intent. This can be a challenging task, as malware is often obfuscated or encrypted to make it difficult to understand. We use ML which is used to generate features that can be used to train machine learning models to detect malware. These features can be based on the code itself, or on the behavior of the malware when it is executed.
[Report]