About Interpolation and Prediction Functions
Interpolation is the process of finding intermediate values in data by fitting appropriate functions (typically polynomials) piece-wise through adjacent data points. This method contrasts with regression, in which a single function is fit, in a least-squares sense, through all data points.
Since interpolation functions must pass through all data points, they are very sensitive to spurious data. If your data is noisy, consider using a regression function instead.