en
0.25
0.5
0.75
1.25
1.5
1.75
2
The influence of weighting the k-occurrences on hubness-aware classification methods
Published on Nov 04, 20112886 Views
Hubness is a phenomenon present in many highdimensional data sets. It is related to the skewness in the distribution of k-occurrences, i.e. occurrences of data points in k-neighbor sets of other data
Related categories
Chapter list
The influence of weighting the k-occurrences on hubness-aware classification methods00:00
Presentation outline00:23
Nearest-neighbor methods in machine learning00:44
The curse of dimensionality and how it affects k-NN methods01:14
Hubs: the influential neighbors02:11
Why it matters03:20
Related work: hubness-aware classification methods04:31
An example: hubness-based weighting04:58
The idea05:26
Our goal06:09
The weighted counts06:13
The data07:23
An observed increase in hubness07:51
Average results (k=30)08:52
So, where does the improvement come from?09:41
The vowel dataset: h-FNN09:56
The conclusion10:52
Thank you for your attention11:14