This work studies the impact of using dynamic information as features in a machine learning algorithm for the prediction task of classifying critically ill patients in two classes ...
Background: Data from metabolomic studies are typically complex and high-dimensional. Principal component analysis (PCA) is currently the most widely used statistical technique fo...
Gift Nyamundanda, Lorraine Brennan, Isobel Claire ...
Abstract. Recent results on robust density-based clustering have indicated that the uncertainty associated with the actual measurements can be exploited to locate objects that are ...
Background: An important challenge for transcript counting methods such as Serial Analysis of Gene Expression (SAGE), "Digital Northern" or Massively Parallel Signature ...
Abstract. An effective way to examine causality is to conduct an experiment with random assignment. However, in many cases it is impossible or too expensive to perform controlled ...