What are economic trends?

What kind of statistical model/machine learning algorithm is used to predict (t+1) Google Trends?

• The old Google Trends had a predictive component.  What kind of statistical model or algorithm was being used to predict the near term trends index?  Asked another way, given you have the past weeks/months of Google Trends data, what is the best way to estimate the next week/months Trends index?  Prediction will never be perfect, but what provides a best fit?

• Answer:

  Google Trend uses various models to predict the near term trends index. It depends a lot on the data which they already have and the earlier used model for the prediction of trends . Google Trends adds it's data along with some constant  αα\alpha which constantly evolves ( learns via self learning neural networks) with time. An example can be predicting some economic index using Google Trends Proposed Model : Fit the best model you can using the data you have (which may often be past values of the time series itself.) Add Google Trends data as an additional predictor . See how the out of sample forecast improves using mean absolute error using a rolling window forecast. Particularly interest in turning points since they are the hardest thing to forecast. Issues with Google Trends : Mixed frequency: Trends is available daily/weekly basis while series of interest may be  weekly or monthly. Google Trends is an index: normalized query share using broad match Must have at least 50 observations to appear in Google Trends due to privacy policy. Google Trends is sampled data, and changes slightly from day to day Can look at session context Google Trend Analysis of relationship between Unemployment and Recession : Here they used the method of autoregressive modelling . The autoregressive model basically specifies that the output variable depends linearly on its own previous values. For long term models  trend + seasonal + residual + Kalman regression  works decently well .Some other factors which can be considered while making a model for prediction are: Causality Counterfactuals Correlation Effects Confounding Variables Regression Discontinuity Selection Bias References : http://www.frbsf.org/economic-research/files/Varian-part_1.pdf, https://web.stanford.edu/class/ee380/Abstracts/140129-slides-Machine-Learning-and-Econometrics.pdf , http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2405804,  http://en.wikipedia.org/wiki/Kalman_filter , http://www.sciencedirect.com/science/article/pii/S0378426612000507 , http://en.wikipedia.org/wiki/Autoregressive_model , http://en.wikipedia.org/wiki/Confounding