A Deep Learning model that predict forecast the power generated by wind turbine in a Wind Energy Power Plant using LSTM (Long Short Term Memory) i.e modified recurrent neural network.

Source code for my research: PM2.5 Density Prediction based on a Two-Stage Rolling Forecast Model using LightGBM

Driver-based rolling revenue forecast on dbt and DuckDB with base, upside, and downside scenarios and a leak-free backtest that measures accuracy honestly: 3.61 percent MAPE, a 42 percent improvement over a seasonal-naive baseline, guarded by dbt tests.

Autoregressive (AR) models with advanced techniques: model selection, diagnostics, structural breaks, rolling forecasts, Fourier seasonality, exogenous variables, business cycle analysis, and benchmarking for economic time series.

Peramalan deret waktu permintaan obat generik di RS Menur Surabaya menggunakan Box-Jenkins ARIMA(2,1,2) di R, dievaluasi melalui Out-of-Sample Backtesting (MAPE 3,11%) dan validasi Rolling Forecast.

Standalone ARIMA forecasting implementation for the log-transformed Sentiment–Volatility Ratio, using UMCSENT and VIXCLS data from FRED.

Standalone AR forecasting workflow for the Sentiment–Volatility Ratio capstone project, using UMCSI and VIX data with expanding-window evaluation.