△=0.04 0 0 0.5 5 △=0.08 0 0 0.5 △=025 0 0 0.5 Histogram method:easy to implement,especially suitable for 1D data. It is more troublesome for high-dimensional case.The number of cells is large, and a lot of data is needed to avoid too few data of some cells e Selection of the size of the cell is a problem.In the histogram method,once the size of Vis determined,the position of V(such as the grid in 1D)is fixed. 15/45
▶ Histogram method: easy to implement, especially suitable for 1D data. ▶ It is more troublesome for high-dimensional case. The number of cells is large, and a lot of data is needed to avoid too few data of some cells. ▶ Selection of the size of the cell is a problem. In the histogram method, once the size of V is determined, the position of V (such as the grid in 1D) is fixed. ▶ Let’s consider the case when the V’s position is not fixed. 15 / 45
Outline (Level 1) Density Estimation Histogram Method Parzen Window k-nearest neighbors Summary of nonparametric models 16/45
Outline (Level 1) 1 Density Estimation 2 Histogram Method 3 Parzen Window 4 k-nearest neighbors 5 Summary of nonparametric models 16 / 45