utubora-katana/Forecasting-inflation-Arima

Forecasting inflation rates in Kenya using Arima model (0,1,0). # Forecasting Inflation Rates in Kenya Using ARIMA(0,1,0) A time series forecasting project completed for BSc Statistics at Egerton University. ## Project Summary Forecasted Kenyan inflation for 5 years using monthly data from Sept 2024 - Nov 2025. Model: ARIMA(0,1,0) using Maximum Likelihood Estimation. ## Key Findings - **Forecast Method**: Random Walk ARIMA(0,1,0) - **Forecast**: Inflation stabilizes at ~5.0% from 2026-2030 - **Model Evaluation**: Mean Absolute Error calculated and visualized ## Skills Demonstrated `R` `forecast` `ggplot2` `tidyr` `Time Series Analysis` `ARIMA` `Data Visualization` `Model Diagnostics` ## Files in this Repository 1. `arima_analysis.R` - Model fitting, 60-month forecast, line plot, annual barplot 2. `model_evaluation.R` - Absolute error barplot and Actual vs Forecast comparison 3. `plots/` - Folder with all generated visualizations ## Visualizations - 5-Year Inflation Forecast Line Chart - Annual Mean Forecasted Inflation Barplot 2026-2030 - Absolute Forecast Error by Month - Actual vs Forecasted vs Error Multi-line Chart