Sharma, Deepshikha
Topological Analysis of Artificial Neural Network - Vol.5(1), Jan-Jun - New Delhi Journals Pub 2019 - 17-21p.
ABSTRACT This paper analyzes the Artificial Neural Network (ANN) over the metric. The dimension, curvature and mappings are the three characteristics of analysis. The impact of ANN in finite and infinite space is also considered. The Euclidean metric is the limit of generalization. The structure of neuron and its mapping are presented over the finite dimension. The discrete and continuous structure of neurons is studied mathematically and its formulation is proposed with minimized errors for the application context. Hence, this topological study opens several doors of application in discrete spaces.
EXTC Engineering
Topological Analysis of Artificial Neural Network - Vol.5(1), Jan-Jun - New Delhi Journals Pub 2019 - 17-21p.
ABSTRACT This paper analyzes the Artificial Neural Network (ANN) over the metric. The dimension, curvature and mappings are the three characteristics of analysis. The impact of ANN in finite and infinite space is also considered. The Euclidean metric is the limit of generalization. The structure of neuron and its mapping are presented over the finite dimension. The discrete and continuous structure of neurons is studied mathematically and its formulation is proposed with minimized errors for the application context. Hence, this topological study opens several doors of application in discrete spaces.
EXTC Engineering