Looking inside the ANN “black box”: classifying individual neurons as outlier detectors

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 analyzing the weights and bias terms. Some authors claim that an ANN trained as a regression device tends to organize itself by specializing some neurons to learn the main relationships embedded in the training set, while other neurons are more concerned with noise. We suggest a rule for identifying the “noise-related” neurons, and we assume that those neurons are activated only when some unusual values are present. We consider those events as candidates to hold an outlier. The speculative nature of this statement has been tested in an experiment summarized in the paper. We used a set of ANNs trained to predict daily precipitation values for a weather station using as input the records obtained from other stations for the same date. The overall procedure was compared within a Monte Carlo framework with a state-of-the-art method for outlier detection. The results show that: a) some evidence confirms the above mentioned assumption about the different roles of the neurons; b) our rule for classifying neurons as related with noise seems reliable; c) ANN-based outlier detection methods based upon our rule outperformed other well established procedures. The use of the ANN as outlier detector does not require further training, and can be easily applied. If the dataset is believed to have outliers, further refinements in the training process might include removing dubious values once detected by the method