Conformal Predictors for Compound Activity Prediction

TL;DR

This paper applies Inductive Mondrian Conformal Predictors to chemoinformatics for predicting compound activity, effectively handling large, high-dimensional, sparse, and imbalanced datasets with various non-conformity measures.

Contribution

It introduces the use of Inductive Mondrian Conformal Predictors for complex chemoinformatics data, demonstrating their flexibility and effectiveness in this domain.

Findings

Effective handling of large, high-dimensional data

Robust performance across different non-conformity measures

Demonstrated flexibility in dealing with class imbalance

Abstract

The paper presents an application of Conformal Predictors to a chemoinformatics problem of identifying activities of chemical compounds. The paper addresses some specific challenges of this domain: a large number of compounds (training examples), high-dimensionality of feature space, sparseness and a strong class imbalance. A variant of conformal predictors called Inductive Mondrian Conformal Predictor is applied to deal with these challenges. Results are presented for several non-conformity measures (NCM) extracted from underlying algorithms and different kernels. A number of performance measures are used in order to demonstrate the flexibility of Inductive Mondrian Conformal Predictors in dealing with such a complex set of data. Keywords: Conformal Prediction, Confidence Estimation, Chemoinformatics, Non-Conformity Measure.