scholarly journals TNT: An Interpretable Tree-Network-Tree Learning Framework using Knowledge Distillation

Entropy ◽  
2020 ◽  
Vol 22 (11) ◽  
pp. 1203
Author(s):  
Jiawei Li ◽  
Yiming Li ◽  
Xingchun Xiang ◽  
Shu-Tao Xia ◽  
Siyi Dong ◽  
...  
Keyword(s):  
Decision Making ◽  
Decision Tree ◽  
Low Frequency ◽  
Test Case ◽  
Tree Model ◽  
Tree Network ◽  
High Performing ◽  
Learning Framework ◽  
Boosted Decision Tree ◽  
Network Tree

Deep Neural Networks (DNNs) usually work in an end-to-end manner. This makes the trained DNNs easy to use, but they remain an ambiguous decision process for every test case. Unfortunately, the interpretability of decisions is crucial in some scenarios, such as medical or financial data mining and decision-making. In this paper, we propose a Tree-Network-Tree (TNT) learning framework for explainable decision-making, where the knowledge is alternately transferred between the tree model and DNNs. Specifically, the proposed TNT learning framework exerts the advantages of different models at different stages: (1) a novel James–Stein Decision Tree (JSDT) is proposed to generate better knowledge representations for DNNs, especially when the input data are in low-frequency or low-quality; (2) the DNNs output high-performing prediction result from the knowledge embedding inputs and behave as a teacher model for the following tree model; and (3) a novel distillable Gradient Boosted Decision Tree (dGBDT) is proposed to learn interpretable trees from the soft labels and make a comparable prediction as DNNs do. Extensive experiments on various machine learning tasks demonstrated the effectiveness of the proposed method.

scholarly journals Decision Tree Pattern Recognition Model for Radio Frequency Interference Suppression in NQR Experiments

Sensors ◽  
2019 ◽  
Vol 19 (14) ◽  
pp. 3153 ◽  
Author(s):  
Ibrahim ◽  
Parrish ◽  
Brown ◽  
McDonald
Keyword(s):  
Decision Tree ◽  
Radio Frequency ◽  
Interference Suppression ◽  
Airport Security ◽  
Test Case ◽  
Radio Frequency Interference ◽  
Automated Identification ◽  
Tree Model ◽  
Radio Frequency Interference Suppression ◽  
Traditional Processing

Radio frequency interference places a major limitation on the in-situ use of unshielded nuclear quadrupole or nuclear magnetic resonance methods in industrial environments for quality control and assurance applications. In this work, we take the detection of contraband in an airport security-type application that is subject to burst mode radio frequency interference as a test case. We show that a machine learning decision tree model is ideally suited to the automated identification of interference bursts, and can be used in support of automated interference suppression algorithms. The usefulness of the data processed additionally by the new algorithm compared to traditional processing is shown in a receiver operating characteristic (ROC) analysis of a validation trial designed to mimic a security contraband detection application. The results show a highly significant increase in the area under the ROC curve from 0.580 to 0.906 for the proper identification of recovered data distorted by interfering bursts.

scholarly journals Identification of Arrhythmia by Using a Decision Tree and Gated Network Fusion Model

2021 ◽  
Vol 2021 ◽  
pp. 1-13
Author(s):  
Peng Lu ◽  
Yabin Zhang ◽  
Bing Zhou ◽  
Hongpo Zhang ◽  
Liwei Chen ◽  
...  
Keyword(s):  
Deep Learning ◽  
Decision Tree ◽  
Frequency Domain ◽  
Weight Control ◽  
High Reliability ◽  
Low Frequency ◽  
Model Parameters ◽  
Fusion Model ◽  
Tree Model ◽  
Time Frequency

In recent years, deep learning (DNN) based methods have made leapfrogging level breakthroughs in detecting cardiac arrhythmias as the cost effectiveness of arithmetic power, and data size has broken through the tipping point. However, the inability of these methods to provide a basis for modeling decisions limits clinicians’ confidence on such methods. In this paper, a Gate Recurrent Unit (GRU) and decision tree fusion model, referred to as (T-GRU), was designed to explore the problem of arrhythmia recognition and to improve the credibility of deep learning methods. The fusion model multipathway processing time-frequency domain featured the introduction of decision tree probability analysis of frequency domain features, the regularization of GRU model parameters and weight control to improve the decision tree model output weights. The MIT-BIH arrhythmia database was used for validation. Results showed that the low-frequency band features dominated the model prediction. The fusion model had an accuracy of 98.31%, sensitivity of 96.85%, specificity of 98.81%, and precision of 96.73%, indicating its high reliability and clinical significance.

scholarly journals Common Metamodel of Questionnaire Model and Decision Tree Model

2018 ◽  
Vol 10 (3) ◽  
pp. 106
Author(s):  
Mirza Suljic ◽  
Edin Osmanbegovic ◽  
Željko Dobrović
Keyword(s):  
Decision Making ◽  
Decision Tree ◽  
Decision Trees ◽  
Decision Tree Model ◽  
Tree Model ◽  
Meta Model ◽  
The Subject ◽  
Decision Tree Method ◽  
Questionnaire Method ◽  
Tree Method

The subject of this paper is metamodeling and its application in the field of scientific research. The main goal is to explore the possibilities of integration of two methods: questionnaires and decision trees. The questionnaire method was established as one of the methods for data collecting, while the decision tree method represents an alternative way of presenting and analyzing decision making situations. These two methods are not completely independent, but on the contrary, there is a strong natural bond between them. Therefore, the result reveals a common meta-model that over common concepts and with the use of metamodeling connects the methods: questionnaires and decision trees. The obtained results can be used to create a CASE tool or create repository that can be suitable for exchange between different systems. The proposed meta-model is not necessarily the final product. It could be further developed by adding more entities that will keep some other data.

scholarly journals A Boosted Decision Tree Model for Predicting Loan Default in P2P Lending Communities

2019 ◽  
Vol 9 (1) ◽  
pp. 1257-1261
Keyword(s):  
Small Business ◽  
Decision Tree ◽  
Decision Tree Classifier ◽  
Tree Model ◽  
Loan Default ◽  
Accuracy Profile ◽  
Default Prediction ◽  
Tree Classifier ◽  
Social Lending ◽  
Boosted Decision Tree

Loan Default Prediction For Social Lending Is An Emerging Area Of Research In Predictive Analytics. The Need For Large Amount Of Data And Few Available Studies In The Current Loan Default Prediction Models For Social Lending Suggest That Other Viable And Easily Implementable Models Should Be Investigated And Developed. In View Of This, This Study Developed A Data Mining Model For Predicting Loan Default Among Social Lending Patrons, Specifically The Small Business Owners, Using Boosted Decision Tree Model. The United States Small Business Administration (Usba) PubliclyAvailable Loan Administration Dataset Of 27 Features And 899164 Data Instances Was Used In 80:20 Ratios For The Training And Testing Of The Model. 16 Data Features Were Finally Used As Predictors After Data Cleaning And Feature Engineering. The Gradient Boosting Decision Tree Classifier Recorded 99% Accuracy Compared To The Basic Decision Tree Classifier Of 98%. The Model Is Further Evaluated With (A) Receiver Operating Characteristics (Roc) And Area Under Curve (Auc), (B) Cumulative Accuracy Profile (Cap), And (C) Cumulative Accuracy Profile (Cap) Under Auc. Each Of These Model Performance Evaluation Metrics, Especially Roc-Auc, Showed The Relationship Between The True Positives And False Positives That Implies The Model Is A Good Fit.

scholarly journals Customer Retention in Banking Sector using Decision Tree-Neuro Based System

2020 ◽  
Vol 9 (3) ◽  
pp. 481-485
Keyword(s):  
Data Mining ◽  
Decision Making ◽  
Decision Tree ◽  
Banking Sector ◽  
Learning Algorithm ◽  
Decision Tree Model ◽  
Tree Model ◽  
Data Mining Techniques ◽  
Individual Decisions ◽  
Bank Marketing

All the bank marketing campaigns mostly deals with large amount of data. when they need to deal with huge electronic data of customers, then it is very difficult to analyze the data manually or by human analyst. Here comes the picture of data mining techniques to deal with the large amount of data and to come up with useful data which helps in decision making process. All the data mining techniques helps in analyzing the data. some of the techniques that can be used for this bank marketing campaigns are naive bayes, logistics regression technique and Decision tree model technique etc. among all these techniques decision Tree model gives the best solution in analyzing the human decisions. Artificial neural networks is a learning algorithm which learns from multiple individual decisions and their judgements, thus aggregates and generalizes the customers decision making knowledge.

scholarly journals Method Based on Floating Car Data and Gradient-Boosted Decision Tree Classification for the Detection of Auxiliary Through Lanes at Intersections

2018 ◽  
Vol 7 (8) ◽  
pp. 317
Author(s):  
Xiaolong Li ◽  
Yuzhen Wu ◽  
Yongbin Tan ◽  
Penggen Cheng ◽  
Jing Wu ◽  
...  
Keyword(s):  
Decision Tree ◽  
Detection Methods ◽  
Tree Model ◽  
Lane Changes ◽  
The Road ◽  
Navigation Data ◽  
Floating Car Data ◽  
Redundant Data ◽  
Boosted Decision Tree ◽  
Using Data

The rapid detection of information on continuously changing intersection auxiliary through lane is a major task of lane-level navigation data updates. However, existing lane number detection methods possess long update cycles and high computational costs. Therefore, this study proposes a novel method based on floating car data (FCD) for the detection of auxiliary through lane changes at road intersections. First, roads near intersections are divided into three sections and the spatial distribution characteristics of the FCD of each section are analyzed. Second, the FCD is preprocessed to obtain a standardized FCD dataset by removing redundant data through an improved amplitude-limiting average filtering method. Third, a basic classifier for the number of lanes is constructed. Fourth, the final number of lanes of the road section is determined by combining the basic classifier and the gradient-boosted decision tree model. Finally, the presence of an auxiliary through lane at the intersection is determined in accordance with the change in the number of intersection lanes. The method was tested using data for a road in Wuchang District, Wuhan City. Experimental results show that this method can rapidly obtain auxiliary through lane information from the FCD and is superior to other classification methods.

scholarly journals Profit Reinvestment: Main Motives Supporting Financial Decisions

e-Finanse ◽  
2019 ◽  
Vol 15 (4) ◽  
pp. 34-43
Author(s):  
Pavla Pokorná ◽  
Jarmila Šebestová
Keyword(s):  
Decision Making ◽  
Pilot Study ◽  
Decision Tree ◽  
Competitive Advantage ◽  
Risk Taking ◽  
Decision Tree Model ◽  
Economic Literacy ◽  
Tree Model ◽  
Financial Decisions ◽  
Specific Factors

AbstractReinvestment decisions are based on basic the economic literacy of entrepreneurs because they do not want to affect future liquidity or development activities. The main goal of the article is to suggest a simple decision tree model to describe profit reinvestments in a general way based on results of a primary pilot study (128 interviews), where reinvestment behaviour is affected by specific factors like risk taking, competitive advantage or business experience. After that a decision-making tree is suggested to explain the process of reinvestment as determined by the manager.

A novel enhanced decision tree model for detecting chronic kidney disease

2021 ◽  
Vol 10 (1) ◽  
Author(s):  
Avijit Kumar Chaudhuri ◽  
Deepankar Sinha ◽  
Dilip K. Banerjee ◽  
Anirban Das
Keyword(s):  
Chronic Kidney Disease ◽  
Kidney Disease ◽  
Decision Tree ◽  
Decision Tree Model ◽  
Tree Model
Sign in / Sign up

Export Citation Format

Share Document