On Thursday 29 January, PhD student Lucyantie Mazalan will present a research seminar on The Effect of Dimensionality on the Effectiveness of Nearest Neighbour Searching Methods in Chemoinformatics Applications.
Nearest neighbour search in chemoinformatics application aims to identify the closest molecule to the query molecule. This involves measurements using different molecular similarity coefficients to the highly dimensional molecular descriptors. However, the performance of nearest neighbour search has been found to decrease as the dimensionality increases. One reason is because, as the dimensionality increases, the ratio of the distance from the query point to the nearest and farthest neighbour approaches unity.
This study will investigate a range of different similarity coefficients on different molecular dimensionality. The results of this analysis will be used to assess the effect of the dimensionality on the effectiveness of nearest neighbour searching methods in chemoinformatics applications.
The seminar takes place in the Information School, Regent Court, RC 231. All are welcome and there is no need to book.