Sponsored by the IEEE Computational Intelligence Society


For contributions to neural networks and unsupervised learning

For automatic analysis of large data masses, the pioneering accomplishments of Erkki Oja in developing artificial neural networks to tackle today’s overwhelming amount of digital data have been critical in extracting reliable and useful information. He introduced a fundamental learning rule that enables a neural network to find the principal components of input data, leading to efficient compression and feature extraction, known as “Oja’s Rule.” Oja improved upon the Independent Component Analysis (ICA) technique with his FastICA method, which he applied to biomedical signal and images and demonstrated that very efficient source separation of artifacts from actual brain signals was possible. His work on self-organizing maps has been used to classify both still pictures and videos and even for indexing nonverbal data.

An IEEE Life Fellow, Oja is a Distinguished Professor Emeritus of Computer Science and Engineering with Aalto University, Espoo, Finland.