Short-term electricity demand forecasting in Southern Thailand using time series analysis and Holt-Winters.
Zero-shot forecasting studio for sales, prices, traffic, and energy.
📊 Forecast daily support incident volumes to enhance resource planning using advanced time series analysis and reliable forecasting models.
📈 Create a simple product sales forecast system in Python with Tkinter, featuring custom calculations and different chart visualizations for data insights.
Forecast any date+value CSV with 8 models scored by walk-forward validation against naive baselines — so you learn whether the forecast beats doing nothing. statsmodels in the browser.
Demand forecasting and reorder point analysis on 540k retail transactions — with a warehouse manager's decision memo, honest baseline comparisons, and every number in the writeup independently verifiable from raw data.
Production-style electricity demand forecasting with rolling backtesting, per-meter model selection, anomaly detection and CI.
Executive-grade climate analytics dashboard with 10-year trend analysis, Holt-Winters forecasting, anomaly detection, and interactive Plotly visualizations for New Delhi, India (2015–2024).
Weekly medication demand forecasting and reorder-point optimization across 8 drug categories (Kaggle Pharma Sales Data, 2014-2019). Python, statsmodels Holt-Winters, ~27% WMAPE.
Smart time-series forecasting for spreadsheets — auto-selects the best method by backtesting, with confidence bands. Free & open-source.
Quarterly Belarus vodka sales forecasting with Holt-Winters, ARIMA, SARIMAX and Prophet using pandas, statsmodels and scikit-learn.
Benchmarks 4 forecasting methods (Seasonal Naive, Holt-Winters, SARIMA, feature-engineered Random Forest) on 8+ years of monthly retail sales, validated on a strict 12-month holdout. Holt-Winters wins, cutting forecast error 33% vs. baseline (MAPE: 10.8% → 7.2%).
Forecasts monthly sales using Holt-Winters Exponential Smoothing on synthetic time series data, with trend/seasonality modeling and MAE evaluation.
Predictive cost intelligence for LLM spend. A zero-dependency TypeScript library and CLI that forecasts future spend with prediction intervals, attributes cost change, detects drift, and prices model swaps.
Productionized demand forecasting: naive/Holt-Winters/LightGBM, rolling-origin backtest (MAPE/sMAPE/MAE/RMSE), deployed forecast API with prediction intervals. ML ~2.2x better than naive.
Time-series forecasting of daily COVID-19 cases using log-scale damped Holt-Winters, benchmarked against a seasonal-naive baseline. Chronological train/test split, MAE/RMSE/MAPE.
Sales forecasting solution using SQL, Holt-Winters and interactive Power BI dashboards.
Real-time anomaly detection and alerting platform for operational telemetry
Multi-KPI US-healthcare forecasting engine on real CDC/NHSN data — SARIMAX+Fourier / Holt-Winters / Prophet, rolling-CV SMAPE selection, deep EDA, calibrated intervals. Interactive Streamlit demo.
Time series analysis and forecasting with statistical models, exponential smoothing, and curve-fitting techniques.
Add a description, image, and links to the holt-winters topic page so that developers can more easily learn about it.
To associate your repository with the holt-winters topic, visit your repo's landing page and select "manage topics."