akshaykapoor347/13-reasons-why-suicide-Timeseries-analysis

We extract data from google trends based on the term suicide and terms related to it from 2017-01-15 to 2017-03-30. We then apply ARIMA model on it to the to forecast the values for the next 19 days. We then compare it with the actual values of the search results to see the accuracy of the forecast. We check for the stationarity of the series and then use acf and pacf to decide p,d and q values of our arima model. We also apply auto arima and compare our forecast results with the real results. We are basically recreating the analysis performed by the John W. Ayers to see whether there was significant increase in the search results related to the suicide query.

R