Mean Shift Mode Seeking
Follow Gaussian-weighted updates toward stationary points of a density estimate.
Free theory articleFollow Gaussian-weighted updates toward stationary points of a density estimate.
Concept & interpretation
Mean shift is an iterative procedure for locating the maxima of a density function given discrete data sampled from that function. It operates by assigning a window around each data point and repeatedly moving the window to the mean of the data points within it, weighted by a kernel function (often a Gaussian). This effectively traces paths toward local regions of highest density (modes).
Experiment
You can verify the calculations by using the interactive module:Mean Shift Mode Seeking
- Set Start to
2. - Set Bandwidth to
1. - Set Iterations to
1. - Observe that the Current location updates to approximately
1.6037, with a Displacement of-0.3963. - Increase the iterations to trace the full path toward the density peak near
1.5.