Artificial Intelligence
Deep learning solutions spanning lending analytics, music cohort segmentation, and computer‑vision models for port‑operations automation. This portfolio showcases practical applications of machine learning and neural networks—including clustering, sequence modeling, and convolutional architectures—to solve real‑world challenges across finance, entertainment, and logistics.
Lending Analytics Developed a high‑accuracy loan‑default prediction model using deep learning techniques in Python, leveraging Keras and TensorFlow. The model incorporated advanced feature engineering, hyperparameter tuning, and robust evaluation methods to support more reliable credit‑risk assessment.
Music Cohort Segmentation Built unsupervised learning pipelines to cluster music into meaningful cohorts using statistical and machine‑learning techniques. This work explored pattern discovery, genre similarity, and audience segmentation through data‑driven clustering.
Computer Vision for Port Operations Designed convolutional neural network (CNN) models to automate key components of port‑operations monitoring. This included developing a custom CNN from scratch using Keras and implementing a lightweight transfer‑learning solution with pretrained architectures to enhance accuracy and reduce training time.
Project Gallery
