Abstract
The paper proposes a 2D-hybrid system of computational intelligence, which is based on the generalized neo-fuzzy neuron. The system is characterised by high approximate abilities, simple computational implementation, and high learning speed. The characteristic property of the proposed system is that on its input the signal is fed not in the traditional vector form, but in the image-matrix form. Such an approach allows getting rid of additional convolution-pooling layers that are used in deep neural networks as an encoder. The main elements of the proposed system are a fuzzified multidimensional bilinear model, additional softmax layer, and multidimensional generalized neo-fuzzy neuron tuning with cross-entropy criterion. Compared to deep neural systems, the proposed matrix neo-fuzzy system contains gradually fewer tuning parameters – synaptic weights. The usage of the time-optimal algorithm for tuning synaptic weights allows implementing learning in an online mode.
Highlights
Neural networks are spread worldwide to solve various tasks within data mining, such as image pattern recognition, the characteristic property of which is the input data that are fed to large-scale processing
The most famous networks are deep convolutional neural networks (CNNs), architecture of which includes relatively independent components: autoencoder – that transforms the initial imagematrix into a vector-signal of relatively low-dimension
It has been designed to solve a pattern-image recognition task without vectorization of the initial data that are represented in matrix form
Summary
Neural networks are spread worldwide to solve various tasks within data mining, such as image pattern recognition, the characteristic property of which is the input data that are fed to large-scale processing. The most famous networks are deep convolutional neural networks (CNNs), architecture of which includes relatively independent components: autoencoder – that transforms the initial imagematrix into a vector-signal of relatively low-dimension After that this vector signal is fed to the input of multilayer fully connected perceptron that solves recognition-classification task. These systems are quite bulky, consist of an enormous number of tuned parameters-synaptic weights, and require a lot of time as well as huge training datasets for learning. It is important to notice that in [15], a multilayer architectures were discussed, where instead of traditional elementary Rosenblatt perceptron [16], neo-fuzzy neurons were used Even though these networks turned out to be faster than traditional CNNs, they still required a huge amount of training data. It is effectively basing on GNFN to introduce into consideration matrix neo-fuzzy system that is purposed to solve pattern-image recognition task
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
