Pytsal

- Add ARIMA model (In preview)
- Fix `csv_loader` bug which failed to detect frequency automatically
- Add new parameter (size) to TimeSeries class
- Setup function now supports custom train-test split ratio
- Add support for RMSE
- Minor bug fixes and performance enhancements

- Add tutorial and examples for forecasting and anomaly detection under **examples folder**
- Use remote dataset from github include of local files
- Add new pytsal logo
- Enhance size of validation plot
- Upgrade anomaly detection module
- Minor bug fixes and performance enhancements

- Add model comparison plot
- Improve scoring functions
- Update brutlag algorithm

- Update example (forecasting and anomaly_detection)
- Fix pypi readme rendering bug

This release add the following:

- 2 Datasets - one for forecasting and other for anomaly detection
- Time series data loaders
- Basic Time series preprocessing
- Holt Winters
- Rich Time series visualization
- Time series validation
- AutoTS feature for holt winter forecasting model