Building interpretable fuzzy models for high dimensional data analysis in cancer diagnosis

Abstract

Background: Analysing gene expression data from microarray technologies is a very important task in biology and medicine, and particularly in cancer diagnosis. Different from most other popular methods in high dimensional bio-medical data analysis, such as microarray gene expression or proteomics mass spectroscopy data analysis, fuzzy rule-based models can not only provide good classification results, but also easily be explained and interpreted in human understandable terms, by using fuzzy rules. However, the advantages offered by fuzzy-based techniques in microarray data analysis have not yet been fully explored in the literature. Although some recently developed fuzzy-based modeling approaches can provide satisfactory classification results, the rule bases generated by most of the reported fuzzy models for gene expression data are still too large to be easily comprehensible.

Results: In this paper, we develop some Multi-Objective Evolutionary Algorithms based Interpretable Fuzzy (MOEAIF) methods for analysing high dimensional bio-medical data sets, such as microarray gene expression data and proteomics mass spectroscopy data. We mainly focus on evaluating our proposed models on microarray gene expression cancer data sets, i.e., the lung cancer data set and the colon cancer data set, but we extend our investigations to other type of cancer data set, such as the ovarian cancer data set. The experimental studies have shown that relatively simple and small fuzzy rule bases, with satisfactory classification performance, can be successfully obtained for challenging microarray gene expression datasets.

Conclusions: We believe that fuzzy-based techniques, and in particular the methods proposed in this paper, can be very useful tools in dealing with high dimensional cancer data. We also argue that the potential of applying fuzzy-based techniques to microarray data analysis need to be further explored.

Figure 1

Membership functions from multiple fuzzy partitions. 15 membership functions from four fuzzy partitions of the domain interval [0, 1]. S, MS, M, ML and L denote Small, Medium Small (relatively small), Medium, Medium Large (relatively large) and Large, respectively. DC denotes “Don’t Care” membership function.

Figure 2

The rule extraction process for the lung cancer data set. Left (UP): The fitness value of the best rule base found in the population; Right (UP): The testing accuracy given by the best rule base; Left (Down): The total number of fuzzy rules in the rule base; Right (Down): The sum of the length of all rules in the rule base.

Figure 3

The rule extraction process for the ovarian cancer data set. Left (UP): The fitness value of the best rule base found in the population; Right (UP): The testing accuracy given by the best rule base; Left (Down): The total number of fuzzy rules in the rule base; Right (Down): The sum of the length of all rules in the rule base.