Time series forecasting with Python -- 3

I would like four models in particular:
- ARIMA
- SVM
- LSTM
- Hybrid (deep learning)

The scope of the project is to predict particle pollution, and specifically PM25 and PM10 for the region of Athens.PM25 and PM25 are fine inhalable particles with diameters that are generally 2.5 and 10 micrometers (and smaller) respectively.

Using the above data, we frame the following forecasting problem: based on the pollution of the previous hours, to predict the pollution in the next hour.

The modelling approach involves statistical, machine learning as well as hybrid to be used. Namely

Statistical model
- ARIMA
- AutoCorrelation Function (ACF)
- Partial AutoCorrelation Function (PACF)
- plot_acf() και plot_pacf() libraries statsmodels
- Machine learning model
- SVM (Support vector machines)
- Neural networks
- LSTM (Long Short-Term Memory)
-Hybrid
- CNN (convolutional neural network) – LSTM(Long Short-Term Memory)
OR
- ARIMA – MLP (multi layer perceptron)

The implementation should be in python using a jupyter notebook.
Thanks,