Research Projects
Developed a comprehensive platform to optimize the process of obtaining birth control and minimize potential side effects. Leveraged Azure AI and ChatGPT-4's retrieval-augmented generation to create a Contraceptive Chatbot Assistant, designed for integration within OBG/YN patient portals.
Technical Details: Combined data from over 200 peer-reviewed articles with differential equations representing typical hormone level fluctuations to build a deep learning algorithm. This algorithm provides personalized recommendations based on user input and medical documents, precisely calculating dosage recommendations based on individual menstrual cycles.
Key Achievements: Presented at the National Conference of Undergraduate Research (NCUR) in Long Beach, CA, spreading awareness about the lack of research in women's health and reproductive rights. The presentation focused on using Generative AI in healthcare, privacy protection practices, and ethical dilemmas, emphasizing the importance of a human-centered approach.
Skills: Project Management, Research, Microsoft Azure, LLM, NLP, Data Mining, Python, Hormone Balancing.
Ethics and Innovations: Navigating the Debate of AI in Healthcare
Chatbot Platform: Hormone Contraceptive Concentration
Jul 2023 - Apr 2024
Authored a comprehensive research paper on the ethical use of Generative AI in healthcare. Explained the basics of ML and current standards. Explored strategies for safeguarding patient data privacy and security for organizations, small businesses and individuals alike.
Skills: Research Ethics, Data Protection, Data Security.
Jul 2023 - Nov 2024
Mind-Controlled Video Game: EMOTIV EEG & BCI
Developed a brain-computer interface to control a ping video game using brainwaves. Programmed the game in Java, linking mental commands to keystrokes and mouse operations via EMOTIV Xavier Control Pane and EmoKey software.
Skills: BCI, Pattern Recognition, Data Analysis, Algorithms, Java, Computer Vision.
Jan 2023 - Jun 2023
Deep Learning for Pneumonia Image Classification
Built a CNN model with 94% accuracy to identify pneumonia from Chest X-Ray images using Python and Keras with TensorFlow as a backend. Cleaned dataset attained from Kaggle.
Skills: CNN, TensorFlow, Pattern Recognition, Neural Networks, Data Analytics.
Dec 2023
Political Bias in Article Classification using k-NN
Using a k-NN algorithm, we built a classifier to detect political bias in articles. Key steps included data preparation, TF-IDF feature extraction, hyperparameter tuning, and classification.
Skills: CNN, TensorFlow, Pattern Recognition, Neural Networks, Data Analytics.
May 2024