Abstract | ||
---|---|---|
In this paper we present an algorithm for finding an optimal configuration of the artificial neural network that is used for the classification in our use case based effort estimation tool. This approach is based on feed-forward artificial neural network and is trained using the back-propagation training algorithm. Our goal is to find the optimal number of hidden neurons and the optimal number of training iterations to be able to reach maximal accuracy of neural network during the estimations. We demonstrate the usage of the proposed algorithm and its result on the estimation example that contains training and testing datasets of UseCases obtained from real software project development. |
Year | DOI | Venue |
---|---|---|
2015 | 10.3233/978-1-61499-611-8-199 | Frontiers in Artificial Intelligence and Applications |
Keywords | Field | DocType |
Neural Networks,Feed-forward,Back-propagation,Softmax,Classification,UseCases | Discrete mathematics,Feedforward neural network,Time delay neural network,Artificial intelligence,Mathematics | Conference |
Volume | ISSN | Citations |
280 | 0922-6389 | 0 |
PageRank | References | Authors |
0.34 | 0 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Radoslav Strba | 1 | 0 | 3.04 |
Jakub Štolfa | 2 | 13 | 10.23 |
Svatopluk Štolfa | 3 | 25 | 13.96 |