Neural networks: a new tool for predicting thrift failures

Article Abstract:

An artificial neural network used in predicting the probability of failure for savings and loan associations is evaluated against a logit model. Results show that the network is more reliable than the logit model in predicting failure outcomes for thrift institutions. This implies that neural networks may serve as models for failure prediction although its inability to provide adequate explanations for the institution's failure may be restrictive.

Evaluation, Savings, Logits

---

Improving decision effectiveness of artificial neural networks: a modified genetic algorithm approach

Article Abstract:

Modified genetic algorithm (MGA), a global search technique, is a training method to improve generalizability and identify required inputs in a neural network (NN) model. Different views, tests and patterns underlying datasets are discussed.

Analysis, Decision support systems, Genetic algorithms, Neural network, Decision support software

---

The effect of misclassification costs on neural network classifiers

Article Abstract:

A new study investigates the cost of misclassification in neural network classifiers.

Research, Classification

Similar abstracts:

• Abstracts: Employee voice: a legal perspective. The effects of participation on B2B exchanges: a resource-based view
• Abstracts: Managing and improving the forecasting process. Foreign currency risk management with probit analysis
• Abstracts: Simultaneous determination of joint product cost allocation and cost-plus prices. Product positioning and pricing under production cost considerations
• Abstracts: The positive effects of easy goals on decision quality and risk propensity in an MCPLP task. Task-structure relationship of information systems development subunit: a congruence perspective

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