High Dimensional Data Mining in Time Series by Reducing Dimensionality and Numerosity