TY - GEN
T1 - On the relation between kNN accuracy and dataset compression level
AU - Blachnik, Marcin
N1 - Publisher Copyright:
© Springer International Publishing Switzerland 2016.
PY - 2016
Y1 - 2016
N2 - The paper discusses how instance selection can be used to asses the kNN performance. There exists a strong correlation between the compression level of the dataset obtained by instance selection methods and the prediction accuracy obtained the k-NN classifier trained on full training dataset. Based on two standard algorithms of instance selection namely CNN and ENN, which belong to two different groups of methods, so called condensation and editing methods, we perform empirical analysis to verify this relation. The obtained results show that this relation is almost linear, so that the level of compression is linearly correlated with the accuracy. In other words by knowing the compression of instance selection methods we are able to estimate the accuracy of the final kNN prediction model.
AB - The paper discusses how instance selection can be used to asses the kNN performance. There exists a strong correlation between the compression level of the dataset obtained by instance selection methods and the prediction accuracy obtained the k-NN classifier trained on full training dataset. Based on two standard algorithms of instance selection namely CNN and ENN, which belong to two different groups of methods, so called condensation and editing methods, we perform empirical analysis to verify this relation. The obtained results show that this relation is almost linear, so that the level of compression is linearly correlated with the accuracy. In other words by knowing the compression of instance selection methods we are able to estimate the accuracy of the final kNN prediction model.
UR - https://www.scopus.com/pages/publications/84976639730
U2 - 10.1007/978-3-319-39378-0_46
DO - 10.1007/978-3-319-39378-0_46
M3 - Conference contribution
AN - SCOPUS:84976639730
SN - 9783319393773
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 541
EP - 551
BT - Artificial Intelligence and Soft Computing - 15th International Conference, ICAISC 2016, Proceedings
A2 - Tadeusiewicz, Ryszard
A2 - Zadeh, Lotfi A.
A2 - Rutkowski, Leszek
A2 - Korytkowski, Marcin
A2 - Scherer, Rafał
A2 - Zurada, Jacek M.
PB - Springer Verlag
T2 - 15th International Conference on Artificial Intelligence and Soft Computing, ICAISC 2016
Y2 - 12 June 2016 through 16 June 2016
ER -