Integrated gene finding: Feed-forward neural networks

• Supervised learning

• Training to discriminate between several feature classes

• Computing units

• Gradient descent optimization

• Multi-layer networks

• Limitations

  • Black-box predictions
  • Local minima

• Example:

  • GRAIL, Uberbacher et al. (1991), PNAS, 88, 11261-11265.

Previous slide

Next slide

Back to first slide

View graphic version