Advanced Time Series Forecasting Toolkit

Develop a comprehensive time series forecasting framework in Python that supports multiple advanced forecasting methodologies including ARIMA, Prophet, exponential smoothing, and machine learning-based approaches. Create a modular system that can automatically select the most appropriate forecasting method, perform hyperparameter tuning, generate confidence intervals, and provide detailed forecast accuracy metrics. Include visualization tools and support for handling complex seasonal and trend patterns.