Broward College

- Analysis of ARIMA(X) and (AR-X) GARCH Models Used for Monthly Consumer Price Index of Energy Services in South Florida (Nov 1977 - Jan 2024)
- Consumer Price Index data sourced from U.S. Bureau of Labor Statistics, accessible via FRED.
- External factors including hurricane events and economic events (e.g. FPL rate changes & COVID pandemic) incorporated in ARIMAX model as exogenous variables.
- Granger Causality Tests to assess causality of hurricane and economic events on CPI.
- Independent T-Tests to compare pre-event and post-event CPI averages.
- Event Study using Generalized Least Squares with Autoregressive Errors (GLSAR) Linear Regression Model to measure specific impacts of events on CPI.
- Seasonal Decomposition of Time Series by Loess (STL) to extract seasonal and trend components.
- Autocorrelation Function (ACF) and Partial Autocorrelation Function (PACF) to identify relevant lags.
- Augmented Dickey-Fuller (ADF) test to check for stationarity.
- Used a First Order Difference Transformation to Achieve Stationarity
- Used Auto-ARIMA and Grid-Search methods to determine optimal ARIMA and ARIMAX model parameters.
- Identified optimal parameters as ARIMA (2, 0, 2) and ARIMAX (2, 0, 2) through Grid-Search.
- Split data into training and testing subsets with 70/30, 80/20, and 90/10 ratios and compared results.
- K-Fold Cross-Validation with TimeSeriesSplit and Blocked Cross-Validation to assess model reliability.
- Implemented GARCH (Generalized AutoRegressive Conditional Heteroskedasticity) and AR-X (eXogenous variables) GARCH models using residuals from ARIMA(X) to handle volatility in residuals from ARIMA(X) indicated by McCleod-Li tests.
- Used Expanding Window and Purged Cross-Validation to evaluate GARCH models' performance under different conditions.

Student Research Intern

Taplio

Chinelo Osuji's Linkedin Analytics

Chinelo Osuji

Want to drive more opportunities from LinkedIn?

Taplio

.css-1jgreyo{background-image:linear-gradient(284deg, rgba(39,43,72,1) 0%, rgba(33,47,183,1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Chinelo Osuji's Linkedin Analytics

Chinelo Osuji

Want to drive more opportunities from LinkedIn?

Chinelo Osuji's Linkedin Analytics