All Relevant Feature Selection Methods and Applications - Feature Selection for Data and Pattern Recognition - page 20

2.3.1 Classification
The results of the classification survey are in general agreement with intuitive expec-
tations, see Fig. 2.2 . Increasing the dimensionality of the problem makes it more
difficult, adding noise to the problem makes it more difficult, and increasing the
number of objects helps in building better models.
Fig. 2.2 Classification results for four variants of hypercube problem. To p 2D and 3D models.
Bottom 4D and 5D models. Labels for series are constructed from the first letter of the series name
and dimension of the problem, for example X3 denotes three dimensional data sets from XOR
series, H5 denotes five dimensional data sets from HYPER series etc