NEC develops automatic optimization technology for deep learning

In an effort to facilitate improvements in recognition accuracy, NEC has developed automatic optimization technology for deep learning. In a statement, the company explains that if deep learning systems become excessively familiar with data, they become unable to accurately recognize data that they have not learned. This "overtraining" results in degradation of recognition accuracy when dealing with data that was not used in the learning process. To prevent overtraining, "regularization" technology is commonly used, which regulates the extent of learning to prevent it from reaching an excessive degree. "This technology predicts the progress of learning at every layer based on the structure of an artificial neural network, and enables regularization to be automatically configured accordingly," said Akio Yamada, General Manager of NEC's Data Science Research Laboratories.