KDE: Kernel Density Estimation
A smooth strongly peaked function - at the position of each data point .
like histogram, bugiven date set and band width makes KDE unique and smooth
- box kernel
- Epanechnikov kernel
- Gaussian kernel
rescale:
http://rightthewaygeek.blogspot.com/2015/09/kernel-density-estimation.html
area under the curve is still one
expected mean-square error
Fourier transform
CDF cumulative distribution
A smooth strongly peaked function - at the position of each data point .
like histogram, bugiven date set and band width makes KDE unique and smooth
- box kernel
- Epanechnikov kernel
- Gaussian kernel
rescale:
http://rightthewaygeek.blogspot.com/2015/09/kernel-density-estimation.html
area under the curve is still one
expected mean-square error
Fourier transform
CDF cumulative distribution
Look at slope. Steep: grow fast
Also, human eye can't compare area under histogram but CDF.***
Tells us what fraction of points fall between any two values.
Probability plot
Summary statistics and box plots
Under certain assumption, mean and SD is useful (Unimodal distribution, single peak)
Comments
Post a Comment