www-ai.cs.tu-dortmund.de/de/LEHRE/FACHPROJEKT/SS12/paper/outlier/marascu2009.pdf
into c).
5. PARAMETERLESS OUTLIER DETECTION
Seq a b c s1 2 0 1 s2 1 0 1 ...
Seq Pn
i=1 similMatrix(s, si) s1 16 s2 14 sn 13 s3 11 ...
sn−1 1
Figure 1: Distances between sequences
Most previous work in outlier [...] values: [1, 1, 2, 5, 9, 10, 13, 15]. Its Haar wavelet transform is illustrated by the following table:
Level Approximations Coefficients 8 1, 1, 2, 5, 9, 10, 13, 15 4 1, 3.5, 9.5, 14 0, -1.5, -0.5, -1 2 2.25 [...] min(s1i , s2i).
For instance, with two sequences s1 and s2 in the matrix of Fig- ure 1, this sum is: s1a + s2b + s2c = 1 + 0 + 1 = 2.
Sometimes, the alignment has to be refreshed and cannot be up- dated …