CGSC170 Lecture 8: CGSC170 Artificial Neural Networks
11 May 2017
School
Department
Course
Professor
Document Summary
Artificial neural network: a network composed of nodes and links with one or more layers. Nodes: (units) the input or outputs in an ann- they function to transmit information throughout the system. Threshold value: if the input into a node is greater than or equal to its threshold value, the node will fire. If the input is below- it will not. Links: the connections among nodes- they allow nodes to send signals among other nodes. Weights: the links have weights that specify the strength of connection among nodes. Input nodes do not distinguish inputs, they simply receive the summed value. Inputs and outputs are not symbolic but they are representational and they do have meaning. Task: design a computerized sonar system to differentiate between sonar echoes returned from mines versus those that are from rocks. Noisy data= data that have irrelevant background information that obscures the signal (size, shape, orientation, distance, variation in material, interference)
