Research on Multi-valued and Multi-labeled Decision Trees

Li, Hong; Zhao, Rui; Chen, Jianer; Xiang, Yang

doi:10.1007/11811305_27

Cited by 4 publications

(3 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…5, 407-420 (2018) tree challenge [11]. A modest number of algorithms have been suggested to construct multi-labelled decision trees [12,13,14,15,16,17]. In these algorithms, various functions in the traditional decision tree algorithms are replaced by functions fit for handling multi-labelled data, primarily based on measures for the similarity between one or more sets.…”

Section: Determine the Decision Treementioning

confidence: 99%

An Approach for Determining Rules used to Select Viable Junction Design Alternatives Based on Multiple Objectives

Bezembinder¹,

Wismans²,

Berkum³

2018

Adv. sci. technol. eng. syst. j.

View full text Add to dashboard Cite

Transport planners and engineers frequently face the challenge to determine the best design for a specific junction. Many road design manuals provide guidelines for the design and evaluation of different junction alternatives, however these mostly refer to specialized software in which the performances of design alternatives can be modelled. In the first stage of the design process, such assessments of many alternatives are undesirable due to time and budget constraints. There is a need for quick design rules which need limited input data. Although some of these rules exist, their usability is limited due to inconsistencies in rules and non-transparency in combination with objectives. In this paper, we present an approach by which consistent and transparent junction design rules can be determined. The resulting rules can be used to predict a set of viable junction design alternatives for the first stage of the junction design assessment process. The predicted set is in fact the Pareto optimal set of solutions for multiple objectives, e.g. regarding operational, safety and/or environmental impact. The Pareto optimal set of solutions always contains the best solution, whatever set of weights is used for different objectives in a later stage of the assessment process, thus handling multiple objectives in a straightforward manner. The rules are derived from a dataset by using decision tree data mining techniques. For this, a large dataset is first generated, using performance models, with Pareto optimal sets of junction design alternatives for a large amount of, randomly generated, traffic volumes. The approach is applied and evaluated on cases for two different countries. Results show that for over 90% of the situations the Pareto optimal set can be predicted by the new rules, whereas existing rules hardly reach 33%. The new rules provide junction design alternatives with a better performances.

show abstract

Section: Determine the Decision Treementioning

confidence: 99%

An Approach for Determining Rules used to Select Viable Junction Design Alternatives Based on Multiple Objectives

Bezembinder¹,

Wismans²,

Berkum³

2018

Adv. sci. technol. eng. syst. j.

View full text Add to dashboard Cite

show abstract

“…Comité expand binary decision tree to process multi-label data [4]. Li Hong proposed a more effective algorithm of multi-label decision tree SCC_SP [5]; they adopt an evaluation method of class-label set similarity by introducing identity and consistency of sets, and compute the class-label set similarity of multi-label data by this way, then regard it as the evaluation index of performance of attribute classification.…”

Section: Introductionmentioning

confidence: 99%

Multi-label Classification based on Association Rules with Application to Scene Classification

et al. 2008

2008 the 9th International Conference for Young Computer Scientists

Self Cite

View full text Add to dashboard Cite

In this paper, a multi-label classification based on association rules is proposed. To deal with multiple class labels problem which is hard to settle by existing methods, this algorithm decomposes multi-label data to mine single-label rules, then combines labels with the same attributes to generate multi-label rules. It extracts partial dataset features to build the initial classifier through assembling, and conducts classification prediction by assembling the classifiers. Thus, the computational complexity caused by the high dimensional attributes decreases while the performance and efficiency increases. Then, the multi-label classification algorithm based on association rules which achieve good performance in an application to scene classification.

show abstract

“…A decision tree classi er named MMC (multi-valued and multi-labelled classi er) was presented by Y. Chen, Hsu and Chou (2003), which was improved in MMDT (multi-valued and multi-labelled decision tree) by Chou and Hsu (2005), in SSC (similarity of same and consistent) by Li, Zhao, Chen and Chang (2006) and in AMDT (a new multivalued and multi-labelled data algorithm) by Yi, Lu and Liu (2011) and Yi, Yan, Lu and Liu (2013). In these algorithms, various functions in the traditional decision tree algorithms are replaced by functions t for handling multi-labelled data, primarily based on measures for the similarity between one or more sets.…”

Section: Multi-label Decision Tree Methodsmentioning

confidence: 99%

Junction design rules

Bezembinder¹

View full text Add to dashboard Cite

It was on a cloudy day in 2011 that a colleague at Windesheim brought the Doctoral Grant for Teachers, issued by the Dutch Research Councel (NWO), to my attention. At that moment, I was rather happy teaching tra c engineering and supervising students performing research on various tra c and transport topics, but had started to get homesick of doing research myself. I contacted Eric van Berkum, who was happy to help me de ne a research proposal and act as supervisor for my PhD-project. I acquired the grant and started my life as a part-time teacher and PhD-student. Combining teaching and research was often satisfying, but sometimes quite a struggle. A University of Applied Sciences o ers a very dynamic work environment, which made me seriously wander from my PhD-path from time to time. Nevertheless, with the help of my supervisors, colleagues, family and friends, I succeeded in nishing this quest. I am proud and happy to present this thesis, which re ects my interests in junction design and tra c and transport modelling.As said, this thesis would not have been completed without the help of many. The rst person I want to thank is my supervisor at the University of Twente, Eric van Berkum. Eric is a very accessible and easy going professor with a truckload of knowledge and experience, and a good sense of humour. I hugely appreciate his patience with me, especially when I repeatedly contacted him to tell him that I would not be able to spend any time on my research in the following weeks or even months. Our talks often started with some nice anecdote or the sports news of that week, before Eric continued asking fundamental questions about the purpose of the analyses and started to o er suggestions for an alternative perspective or additional analyses. It happened more than once, that after a few weeks, although I initially rejected a suggestion, discovered that it was a 'damn good idea'. In the end, Eric and Luc were able to interpret the tone of my 'hmm', whether I agreed with their comments and was inclined to do something with it. That was a crucial incentive to nally nish my project.That brings me to the second person I want to thank, being my co-supervisor, Luc Wismans. I must be careful not to praise him too much, since he will probably remind me of it. Luc's input was very valuable to me. I used his PhD-research and thesis as a template for my own. Luc was always available and willing to answer questions and very generous with his comments in draft versions of papers and thesis-chapters. I often suspected that he somehow managed to put more than twenty-four hours in his days. The only time Luc declined a meeting request was when I o ered to cook a spicy Thai dinner when we had to discuss a couple of thesis-chapters. I would advise anyone to ask Luc to be his or her (co-)supervisor.This research and thesis would not have been possible without the help of my colleagues at Windesheim. Above all, I want to thank my colleagues at the team of Spatial Planning and Mobility. Janet, Marcus, Natascha, Je rey, Hugo and A...

show abstract

Research on Multi-valued and Multi-labeled Decision Trees

Cited by 4 publications

References 7 publications

An Approach for Determining Rules used to Select Viable Junction Design Alternatives Based on Multiple Objectives

An Approach for Determining Rules used to Select Viable Junction Design Alternatives Based on Multiple Objectives

Multi-label Classification based on Association Rules with Application to Scene Classification

Junction design rules

Contact Info

Product

Resources

About