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.

Classification method and device of image, storage medium and electronic equipment

Abstract: The embodiment of the application discloses a classification method and device of an image, a storage medium and electronic equipment. The pushing method includes: when an operation of moving an imageto a classification library is detected, collecting a multi-dimensional feature, which corresponds to the image, to use the same as a sample, and constructing a sample set corresponding to multiple classification libraries; carrying out sample classification on the sample set according to information gain of features on sample classification to construct a decision tree model of the classification libraries, wherein output of the decision tree model is the multiple corresponding classification libraries; collecting a multi-dimensional feature, which corresponds to the to-be-classified image,to use the same as a prediction sample when an image classification instruction is detected; and predicting a corresponding classification library according to the prediction sample and the decision tree model. Therefore, intelligent classification of the image is realized, and a classification accuracy rate of the image is increased.

Information geometry of the family of Markov kernels defined by a context tree

Abstract: We prove that a tree model is an exponential family (e-family) of Markov kernels, if and only if it is an FSMX model. The notion of e-family of Markov kernels was first introduced by Nakagawa and Kanaya ('93) in the one-dimensional case. Then, Nagaoka ('05) gave its established form, and Hayashi & Watanabe ('16) discussed it. A tree model is the Markov model defined by a context tree. It is noted by Weinberger et al., ('95) that tree models are classified into two classes; FSMX models and non-FSMX models, depending on the shape of their context trees. The FSMX model is a tree model and a finite state machine. We further show that, for Markov models, the e-family of Markov kernels is equivalent to the asymptotic e-family, which was introduced by Takeuchi & Barron ('98). Note that Takeuchi & Kawabata ('07) proved that non-FSMX tree models are not asymptotic e-families for the binary alphabet case. This paper enhances their result and reveals the information geometrical properties of tree models.

A Result Verification of Decision Tree Model for Industrial Wireless Sensors Selection using Analytic Hierarchy Process

Abstract: Nowadays, an industrial wireless sensor (IWS) is interested and widely used in a Gas industry in Thailand. There are many vendors trying to improve quality of IWS products for gaining advantage in competitive market. Therefore, choosing the IWS becomes a challenge for users not only the brand name and price but also several factors needed to be considered, e.g., data rate, output power, operating voltage, transmitting current, receiving current, and operating temperature. Selecting the proper IWS is considered as a multi-objective decision problem that is complicated for an engineer and a project manager. The classification method using a Decision Tree)DT(model can be applied to solve such the problem, but the accuracy is depended on the number of historical data. For IWS in the Gas industry in Thailand, not only the number of training and testing data is limited but also there are only a few brands that are chosen regularly. In this paper, the method of applying the DT model for IWS classification and selection is presented. Then, the classification result of the DT model is verified by using an Analytic Hierarchy Process)AHP(for confirming whether it is accurate based on the limit number of historical data. The verification result can be preliminarily ensured that the DT model can be applied as a decision tool for choosing the appropriate IWS accurately.

Decision Support Solution To Business Failure Prediction

Abstract: This paper aims to develop a practical decision support solution to business failure prediction, as early warning signals of potential financial distress could become a true asset in the decision making process of a firm. Several prediction models, such as decision trees and neural networks are built on a sample of Romanian firms and tested for their prediction ability. In order to try to improve the prediction ability of the tree model, we propose a method based on principal component analysis. The high prediction accuracy of the models suggests that the proposed decision support solution can become a practical tool for any decision maker.

A robot loop detection method and device base on deep metric learning combined with bag-of-words tree model

Abstract: The invention discloses a robot loop detection method and device based on deep metric learning combined with a bag-of-words tree model. The method comprises the following steps: 1) inputting a scene video stream whose environment appearance changes for a long time; 2) training and learning that feature extraction network by using a deep metric learning framework; 3) extracting features from that train video stream image by using the feature extraction network; 4) iteratively clustering the obtained features to establish a bag-of-words tree model; 5) inputting the current key frames in the video stream obtain by the robot in real time in the positioning and navigation process of the actual robot; 6) utilizing the feature extraction network to extract the features of the current key frame; 7) adding the features of the current key frame to the bag-of-words tree model; 8) searching and matching image frames with similar image features by the bag-of-words tree model, and carrying out similarity measurement to judge whether the robot meets a loop. The invention can realize high-efficient loop detection in the long-time positioning and navigation process of the robot in a dynamic environment.