Learning Enhancement of CNNs via Separation Index Maximizing at the First Convolutional Layer
This is an incremental improvement for CNN users in classification tasks, as it modifies training without altering network architecture.
The paper tackles the problem of improving CNN performance for classification by optimizing the first convolutional layer to maximize a Separation Index (SI) measure, resulting in demonstrated enhancement across most tested CNNs and datasets.
In this paper, a straightforward enhancement learning algorithm based on Separation Index (SI) concept is proposed for Convolutional Neural Networks (CNNs). At first, the SI as a supervised complexity measure is explained its usage in better learning of CNNs for classification problems illustrate. Then, a learning strategy proposes through which the first layer of a CNN is optimized by maximizing the SI, and the further layers are trained through the backpropagation algorithm to learn further layers. In order to maximize the SI at the first layer, A variant of ranking loss is optimized by using the quasi least square error technique. Applying such a learning strategy to some known CNNs and datasets, its enhancement impact in almost all cases is demonstrated.