Methodology for Forecast Talent pool:
Two different methods are used based on data availability.
- Method 1: Used for cities where historical growth rates were available
- Method 2: Used for cities where only current year growth rate was available.
Extrapolation of historical data: In this technique, growth rate is forecasted using combination of log/linear/exponential/polynomial (2nd degree) extrapolation of historical growth rates. Each extrapolation is analyzed individually as well as in combination with other extrapolations and the best curve is chosen by analyst perception. To see equations for extrapolation refer to last section of this documentation.
Cities where historical growth rates are not available. In this case, all the cities were divided into few profiles, and for each profile a forecast curve was created by method 1 using one of cities for which data is available. This curve was superimposed on current year growth rate of the city to generate forecast.
There are many mathematical models for forecasting trends and cycles. The appropriate model for a particular forecasting application depends on the historical data.