An aspiring Software Developer, Machine learning and Deep Learning enthusiast.

🎓 Education

I am currently enrolled in a master's program in Computer and Information Sciences at Clemson University. Throughout my academic journey, I am actively broadening my understanding and proficiency in the domains of machine learning and deep learning. Simultaneously, I am honing my skills in software development and gaining expertise in cloud technologies.

The ultimate goal of acquiring this diverse set of knowledge and skills is to pave the way for a successful career as a software developer. I aspire to apply my expertise to real-world projects, utilizing my understanding of machine learning, deep learning, software development, and cloud technologies to contribute effectively to the ever-evolving landscape of technology.

By combining theoretical knowledge gained through academic pursuits with practical skills developed through hands-on experiences, I aim to make meaningful contributions to the field of software development. This comprehensive approach not only prepares me for the challenges of the industry but also positions me to excel in dynamic and cutting-edge projects.

🔭 Projects Overview

As a budding enthusiast in the field of deep learning, I am actively engaged in computer vision projects to gain a profound understanding of the intricacies of deep learning and machine learning models. These projects not only allow me to delve deeply into the theoretical aspects but also offer valuable hands-on experience in solving real-world problems. The range of models I've explored thus far includes:

• Feed Forward Neural Networks: I have delved into the fundamentals of neural networks, understanding their architecture and training mechanisms to grasp the core concepts of deep learning.

• Convolutional Neural Networks (CNNs): I've applied CNNs not only to 2D data but also extended my exploration to 3D data. This has provided me with insights into how these specialized neural networks excel in tasks related to image and spatial data.

• Support Vector Machines (SVMs): I've worked with SVMs, comprehending their ability to classify data points by finding the optimal hyperplane. This has broadened my understanding of both linear and non-linear classification.

• Random Forest Classifier: Exploring ensemble methods, particularly the Random Forest algorithm, has given me insights into leveraging the power of multiple decision trees for improved model performance and robustness.

• Long Short-Term Memory (LSTM): Venturing into recurrent neural networks, specifically LSTMs, has enabled me to address sequential data and grasp the temporal dependencies inherent in various applications.

• XGBoosting Algorithm: I've applied the XGBoost algorithm, a powerful gradient boosting technique, to enhance predictive accuracy and handle complex relationships within the data.

Engaging in these diverse models has not only deepened my understanding of the theoretical foundations of machine learning but has also equipped me with the practical skills needed to tackle real-world challenges in computer vision. This hands-on experience is crucial in bridging the gap between theoretical knowledge and its application in solving complex problems.

♾️ Projects List

• Driver Drowsiness Detection - Developed a CNN + LSTM-based deep learning model for real-time analysis of driver behavior in video data, implementing an effective drowsiness detection system. Enhanced road safety by proactively alerting drowsy drivers, mitigating the risk of accidents.
• Quality-Evaluation-of-Skull-Stripped-Brain-MRI-Images - Developed an innovative tool leveraging deep learning (CNN) to automate quality assessment of skull-stripped brain MRI images, minimizing human intervention. Significantly enhanced efficiency and accuracy by 30% in image evaluation within a medical imaging context.
• Exploring SIFT - Explored the concept of Scale Invariant Feature Transform(SIFT) algorithm in OpenCV.
• Semantic Search Engine - A search engine build using AI technology to search for the text documents relevant to the input query. It uses sentence transformer model to encode the text documents. The embeddings of text documents are stored in vector database (Pinecone in this case) such that it is easy to find the relevant documents.
• Determining early readmission of diabetic patients within 30 days of discharge - Engineered a cutting-edge tool using deep learning (CNN) to determine if a diabetic patient gets readmitted within thirty days of discharge. Integrating these findings into clinical workflows, will optimize patient care and elevates healthcare standards with improved treatment outcomes.
• Antibackdoor Learning - Explored a robust training methodology which helps the deep learning models to get trained robustly on poisoned data sets.

🍀 Want to get to know my skills ?

💬 Ask me about - C, CPP, Python, Java, HTLM, CSS, Git, Scikit-learn, Numpy, Pandas, Pytorch, AWS.

• Please visit my portfolio to download my resume and CV - https://lakshmigayathri19.github.io/Portfolio/.

📫 How to reach me ?

📩 Email: [email protected]

• Feel free to connect with me on LinkedIn