A neural network approach to solve the stable matching problem

Article Abstract:

A neural network approach that dynamically generates solutions to stable matching problems is introduced. It makes use of two types of neurons, namely, the maximum selection neuron and the maximum cut-off neuron, to construct a matching neural net that represents the game theory problem and derives solutions in a distributed parallel processing environment. In addition to computational efficiency, the neural net converges to a steady state always despite the lack of network changes.

Game theory, Matching theory

---

Using neural networks for trauma outcome evaluation

Article Abstract:

The viability of artificial neural network models for forecasting the mortality of trauma patients was investigated. Specifically, neural networks trained using the backpropagation and quickprop techniques were used to assess the physiologic and anatomic injury scores of trauma patients with penetrating injuries. The results demonstrated the dependence of classification accuracy on both the training set and testing set composition.

Pattern recognition, Pattern recognition (Computers)

---

Similar abstracts:

• Abstracts: A neural network model for scheduling problems. A comparison of neural networks and linear scoring models in the credit union environment
• Abstracts: Comparative performance of the FSCL neural net and the K-means algorithm for market segmentation. An adaptive, multiple restarts neural network algorithm for graph coloring

This website is not affiliated with document authors or copyright owners. This page is provided for informational purposes only. Unintentional errors are possible.