Handling uncertainties in toxicity modelling using a fuzzy filter

Abstract

A fundamental concern in the Quantitative Structure-Activity Relationship approach to toxicity evaluation is the generalization of the model over a wide range of compounds. The data driven modelling of toxicity, due to the complex and ill-defined nature of eco-toxicological systems, is an uncertain process. The development of a toxicity predicting model without considering uncertainties may produce a model with a low generalization performance. This study presents a novel approach to toxicity modelling that handles the involved uncertainties using a fuzzy filter, and thus improves the generalization capability of the model. The method is illustrated by considering a data set dealing with the fathead minnow (Pimephales promelas) toxicity of 568 organic compounds.

A fuzzy system for modeling the structure-activity relationships in presence of uncertainties

“…The fuzzy filter of [1], [2] maps the input descriptors values (denoted by a vector x = [d 1 d 2 · · · d n ] T ∈ X) to the "filtered" output activity value (denoted by a scalar y f ) with K number of rules of the following type If x belongs to a cluster having centre c 1 then y f = α 1 . .…”

“…the mathematical mapping between input x and filtered output y f is free from uncertainties. Based on some clustering criteria, it was shown in [1] that…”

“…To model the structure-activity relationships, we extend the rule-base (1) of the fuzzy filter as follows If x belongs to a cluster having centre c 1 then y f = α 1 and for an i th −type structure the activitŷ…”

“…We have recently proposed in [1] and [2] the use of a fuzzy filter, designed via some robust identification procedure, to filter out the uncertainties from the modeling problem for improving the generalization performance of the models. This manuscript presents a fuzzy based method of modeling in presence of uncertainties based on following ideas:…”

On the Estimation of Parameters of Takagi–Sugeno Fuzzy Filters

“…The design of the fuzzy filters is an important issue since many real-world applications (e.g., in chemistry [1], biomedical engineering [2], [3], etc.) require the filtering of uncertainties from the experimental data.…”