Current Slide

Small screen detected. You are viewing the mobile version of SlideWiki. If you wish to edit slides you will need to use a larger device.

Efficiency and Interpretability

  • Efficiency of backpropagation: Each epoch (one iteration through the training set) takes O(|D| * w), with |D| tuples and w weights, but # of epochs can be exponential to n, the number of inputs, in worst case
  • For easier comprehension: Rule extraction by network pruning
    • Simplify the network structure by removing weighted links that have the least effect on the trained network
    • Then perform link, unit, or activation value clustering
    • The set of input and activation values are studied to derive rules describing the relationship between the input and hidden unit layers
  • Sensitivity analysis: assess the impact that a given input variable has on a network output. The knowledge gained from this analysis can be represented in rules

Speaker notes:

Content Tools

Sources

There are currently no sources for this slide.