Investigating the Role of Individual Neurons as Outlier Detectors

Abstract

The main body of the literature states that Artificial Neural Networks must be regarded as a "black box" without further interpretation due to the inherent difficulties for analyze the weights and bias terms. Some authors claim that ANN trained as a regression device tend to organize itself by specializing some neurons to learn the main relationships embedded in the training set, while other neurons are more concerned with the noise. We suggest here a rule to identify the "noise-related" neurons in multilayer perceptron ANN, and we assume that those neurons are activated only when some unusual values (or combination of values) are present. We consider those events as candidates to hold an outlier. The use of the ANN as outlier detector does not require further training, and can be easily applied.