O Boz - Proceedings of the eighth ACM SIGKDD international …, 2002 - dl.acm.org
… extracting DecisionTrees from trained Feedforward NeuralNetworks. We show that the new method is able to create DecisionTrees that are close in accuracy to the NeuralNetworks …
… of the knowledge stored in neuralnetworks. In this paper, we … of neuralnetworks and the comprehensibility of decisiontrees… generating decisiontrees that utilizes trained neuralnetworks…
… decisiontree (DT) baseline we set two of the key hyperparameters criterion as ‘gini’ and splitter as ‘best’. For neuralnetwork … of DNDT, decisiontree, and neuralnetwork models on each …
Y Bouzida, F Cuppens - IEEE/IST workshop on monitoring …, 2006 - yacine.bouzida.free.fr
… Section III briefly presents background of neuralnetworks and an improvement of … neural network and its enhancement for new attacks detection. In Section IV, we present decisiontrees …
R Balestriero - arXiv preprint arXiv:1702.07360, 2017 - arxiv.org
… Abstract—In this paper we propose a synergistic melting of neuralnetworks and decision trees (DT) we call neuraldecisiontrees (NDT). NDT is an architecture a la decisiontree where …
GPJ Schmitz, C Aldrich… - … on Neural Networks, 1999 - ieeexplore.ieee.org
… The ANN-DT (artificial neural-networkdecisiontree) algorithm described in this paper can be seen as pedagogical, although it is not bound by the limitations of the above strategies. …
C Jacobsen, U Zscherpel, P Perner - … Machine Learning and Data Mining in …, 1999 - Springer
… Then, we explain what kind of neuralnets and induction of decisiontrees were used for … performance of neuralnets is not significant better than the performance of decisiontrees if we …
IK Sethi - Proceedings of the IEEE, 1990 - ieeexplore.ieee.org
… neuralnetwork. The objective of this paper is to show how this mapping of decisiontrees into a multilayer neuralnetwork … class of layered neuralnetworks, called entropy nets, that have …
… mapping from decisiontrees to deep neuralnetworks. The mapping produces a network with a … per hidden layer, and a set of initial weights that reflect the decisiontree structure. The …