Data Analytics
more coming soon
Time Series Forecasting & Anomaly Detection
Completed a project on AEP Hourly Energy Consumption forecasting and identifying anomalies. The work involved data preparation, change point detection, out-of-time model evaluation, and diagnostics across all periods. Hyperparameter tuning was performed to optimize model performance and improve forecasting accuracy.
Fraud Detection for Credit Card Transactions using PCA & iForest
Developed an anomaly detection modeling pipeline for credit card transactions using PCA and Isolation Forest. Engineered features from raw transaction data through aggregation methods to capture behavioral and temporal patterns. Applied model aggregation techniques to enhance prediction stability and ensure robust fraud detection.
Fraud Detection for Hospital Charges using Autoencoders
Engineered features from raw hospital charge data using aggregation methods to capture behavioral and temporal patterns. Developed a fraud detection pipeline using autoencoders to identify anomalous transactions. Applied model aggregation techniques to improve prediction stability and ensure robust fraud detection.
