Topic
Decision tree model
About: Decision tree model is a research topic. Over the lifetime, 2256 publications have been published within this topic receiving 38142 citations.
Papers published on a yearly basis
Papers
More filters
••
TL;DR: The notion of admissible models are defined as a function of problem complexity, the number of data pointsN, and prior belief to derive general bounds relating classifier complexity with data-dependent parameters such as sample size, class entropy and the optimal Bayes error rate.
Abstract: In this paper we investigate the application of stochastic complexity theory to classification problems. In particular, we define the notion of admissible models as a function of problem complexity, the number of data pointsN, and prior belief. This allows us to derive general bounds relating classifier complexity with data-dependent parameters such as sample size, class entropy and the optimal Bayes error rate. We discuss the application of these results to a variety of problems, including decision tree classifiers, Markov models for image segmentation, and feedforward multilayer neural network classifiers.
3 citations
••
29 Sep 2009TL;DR: The flexible neural tree model is applied for forecasting the housing price index (HPI) and the optimal structure is developed using the Modified Breeder Genetic Programming and the free parameters encoded in the optimal tree are optimized by the Particle Swarm Optimization (PSO).
Abstract: Since the subprime crisis, the variance of housing price is receiving increasing attention especially because of its complexity and practical applications. This paper applies the flexible neural tree model for forecasting the housing price index (HPI). The optimal structure is developed using the Modified Breeder Genetic Programming (MBGP) and the free parameters encoded in the optimal tree are optimized by the Particle Swarm Optimization (PSO), and a new fitness function based on error and Occam's razor is used for for balancing of accuracy and parsimony of evolved structures. Based on the HPI of Shandong province, the performance and efficiency of the applied model are evaluated and compared with the classical multi-layer feed-forward network (MLFN) and support vector machine (SVM) models.
3 citations
•
22 Jan 2014
TL;DR: In this article, a P2P protocol identification method based on a secondary decision tree is proposed, which consists of the following steps: S1, obtaining pure P 2 P protocol flow and non-P 2 P 2 protocol flow, and extracting a primary network flow statistics characteristic set in a specific format; S2, according to the extracted primary network flows characteristic set, respectively training a primary decision tree model set and a secondary Decision Tree model set; S3, extracting from a network the primary networks flow statistics characteristics of a network flow set which accords with a specific trigger
Abstract: The invention discloses a P2P protocol identification method based on a secondary decision tree. The method comprises the following steps: S1, obtaining pure P2P protocol flow and non-P2P protocol flow, and extracting a primary network flow statistics characteristic set in a specific format; S2, according to the extracted primary network flow statistics characteristic set, respectively training a primary decision tree model set and a secondary decision tree model set; S3, extracting from a network the primary network flow statistics characteristics of a network flow set which accords with a specific trigger rule and is provided with a network flow five-element group comprising an IP address to be detected; and S4, identifying the P2P protocol in background flow by using the primary and secondary decision tree model sets obtained in step 2 and the primary network flow statistics characteristics extracted in step 3. By using the method provided by the invention, the severe false reporting caused by using a conventional P2P protocol identification method can be effectively improved, and technical support can be provided for the design and realization of the high-performance flow classification system and the content monitoring system in a high-speed network.
3 citations
••
01 Jan 2009TL;DR: This book describes and analyze algorithms on digraphs and focuses more on graph-theoretical aspects of these algorithms than on their actual implementation on a computer.
Abstract: In this book we often describe and analyze algorithms on digraphs. We concentrate more on graph-theoretical aspects of these algorithms than on their actual implementation on a computer. Thus, in many cases only the most basic knowledge on algorithms and complexity is required and many readers are familiar with it. However, sometimes we use less familiar terminology and notation. In particular, we sometimes say that some problem is fixed-parameter tractable or W[1]-hard.
3 citations