Abstract:In this paper we introduce a new method which employs the concept of "Orientation Vectors" to train a feed forward neural network and suitable for problems where large dimensions are involved and the clusters are characteristically sparse. The new method is not NP hard as the problem size increases. We `derive' the method by starting from Kolmogrov's method and then relax some of the stringent conditions. We show for most classification problems three layers are sufficient and the network size depends on the number of clusters. We prove as the number of clusters increase from N to N+dN the number of processing elements in the first layer only increases by d(logN), and are proportional to the number of classes, and the method is not NP hard. Many examples are solved to demonstrate that the method of Orientation Vectors requires much less computational effort than Radial Basis Function methods and other techniques wherein distance computations are required, in fact the present method increases logarithmically with problem size compared to the Radial Basis Function method and the other methods which depend on distance computations e.g statistical methods where probabilistic distances are calculated. A practical method of applying the concept of Occum's razor to choose between two architectures which solve the same classification problem has been described. The ramifications of the above findings on the field of Deep Learning have also been briefly investigated and we have found that it directly leads to the existence of certain types of NN architectures which can be used as a "mapping engine", which has the property of "invertibility", thus improving the prospect of their deployment for solving problems involving Deep Learning and hierarchical classification. The latter possibility has a lot of future scope in the areas of machine learning and cloud computing.

Research on Forward Neural Network for Corner Classification Based on Real Vectors

Research on Forward Neural Network for Corner Classification

Behavior of CC4 Neural Network Based on Deriving Vector

Textcc: New Feed Forward Neural Network for Classifying Documents Instantly

A document classification approach by GA feature extraction based corner classification neural network

An Extended Corner Classification Neural Network Based Document Classification Approach

Corner2Net: Detecting Objects as Cascade Corners

Research on the Visual Image Processing Based on Neural Network

APPLICATION OF NEURAL NETWORK IN IMAGE PROCESSING

Web document classifier design based on fast classification network

Document Classification With Cc4 Neural Network

A Forward Neural Network Based Relevance Feedback Algorithm Design in Image Retrieval

Document classification approach by rough-set-based corner classification neural network

Remote Sensing Image Classification Based on Neural Networks Designed Using an Efficient Neural Architecture Search Methodology

Some Theorems for Feed Forward Neural Networks

Generalized Learning Vector Quantization for Classification in Randomized Neural Networks and Hyperdimensional Computing

Internet Traffic Classification Using Feed-Forward Neural Network

Research on stereographic projection and it's application on feed forward neural network

Research of Assembling Optimized Classification Algorithm by Neural Network Based on Ordinary Least Squares (OLS).

Social network based link correlation using graph neural network with deep learning architectures for feature vectors prediction and classification

A new genetic algorithm based evolutionary neural architecture search for image classification