As far as in the blog, I have given the explanation and an overview of the Time Series Analysis.
Now I am going to perform the time series on the Logan Passengers and Logan International Flights variables and try to build the predictive models and will find out forecasting with respect to the analysis.
The normal flow I am going to follow for the time series approach are:
- Identify the component.
- Check the time series is stationary or not by using Augmented-Dickey-Fuller (ADF) or Kwiatkowski–Phillips–Schmidt–Shin (KPSS) test.
- If the time series is not stationary, make it by differencing or transformation.
- Pre-work for AR and MA model by using Auto-Correlation Function (ACF).
- Start building the model for AR and MA.
- At the end, apply AR+I+MA.
1. First I have identified my components and variables on which I am going to perform the time series analysis. I have chosen the Logan Passengers and Logan International Flights variables and will try to analyse the correlation in between them and what will be the scope in future according to the number of flights departed from Boston Logan International Airport.
2. After that we shall need to check whether the data is stationary or not, for this I have applied the Augmented-Dickey-Fuller (ADF) test. By this we can check the behaviour of the data and get the stats for the data like p-value, ADF Statistics , and Critical Values. By this technique and after doing the ADF test I got the p-value greater than the 0.05 which means that the time series is not stationary and failed to reject the null hypothesis.
3. After this we need to differencing to make the time series stationary and for this I applied the differencing technique and later I got the p-value around 1.7538458615299656e-10. This shows that the time series is now stationary and rejects the null hypothesis.
At this point, I made my time series analysis stationary and now I am ready to work on my model. In the next blog, I will try to gather some more information about building the model in time series and also find the outcomes of the SARIMAX.