2 Hits in 4.1 sec

Compounds Activity Prediction in Large Imbalanced Datasets with Substructural Relations Fingerprint and EEM

Wojciech Marian Czarnecki, Krzysztof Rataj
2015 2015 IEEE Trustcom/BigDataSE/ISPA  
It is a consequence of huge classes disproportion (even 1000:1), large datasets (over 100,000 of samples) and restricted data representation (mostly high-dimensional, sparse, binary vectors).  ...  In this paper, we try to tackle this problem through three important innovations. First we represent compounds with 2-dimensional, graph representation.  ...  We showed a proof of concept solution for the classification of large dataset of extremely imbalanced compounds.  ... 
doi:10.1109/trustcom.2015.581 dblp:conf/trustcom/CzarneckiR15 fatcat:7jghn6lkljgypgif7uedmqq64y

Applications of Support Vector Machines in Chemistry [chapter]

Ovidiu Ivanciuc
2007 Reviews in computational chemistry  
and 15 non-active compounds) and pIC 50 ¼ 6.0 (28 active and 45 non-active compounds).  ...  ), whereas drugs with no reported cases of TdP in humans were selected as the non-active compounds (204 for training and 39 for prediction).  ... 
doi:10.1002/9780470116449.ch6 fatcat:aumcn53nvfhhhocxvav32rhwzm