Tree retraining in the decision tree learning algorithm

Mitrofanov, S. A.; Semenkin, ‪Eugene

doi:10.1088/1757-899x/1047/1/012082

Cited by 11 publications

(6 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recognition of the urban landscape. The paper compares the best results of the standard compositions considered in (Mitrofanov & Semenkin, 2021) with the proposed approach (EGP).…”

Section: Resultsmentioning

confidence: 99%

Hybrid Evolutionary Approach to Decision Trees Ensembles Design

Mitrofanov¹

2023

European Proceedings of Computers and Technology

View full text Add to dashboard Cite

Decision trees are an efficient data analysis tool. Ensembling methods have been developed on the basis of decision trees. These methods make it possible to obtain a data analysis tool in the form of a composition of trees. The paper proposes a new approach since the development of compositions based on decision trees is an urgent problem. The paper proposes a new hybrid approach to designing the composition of decision trees. The approach is based on the idea of the decision tree application built by a genetic programming algorithm as a technique to determine a machine learning method for object classification. Thus, with the help of the proposed approach the authors carry out a hybridization of a self-configuring genetic programming algorithm and a decision tree. The paper treats decision trees built by a modified algorithm with differential evolution considered as data analysis methods that make decisions concerning a sample objects classification. The proposed method is studied on some classification problems with different types of data and dimensions. The comparison with other methods for building compositions of decision trees is made.

show abstract

“…Recognition of the urban landscape. The paper compares the best results of the standard compositions considered in (Mitrofanov & Semenkin, 2021) with the proposed approach (EGP).…”

Section: Resultsmentioning

confidence: 99%

Hybrid Evolutionary Approach to Decision Trees Ensembles Design

Mitrofanov¹

2023

European Proceedings of Computers and Technology

View full text Add to dashboard Cite

show abstract

“…Various filtering methods have been previously studied and tested to select a separation feature [8]. The results of practical tests showed that the most suitable methods may differ depending on the specific application.…”

Section: Two-level Optimization Approachmentioning

confidence: 99%

“…During the process of creating a decision tree, various problems arise, each of which must be solved. One of the main problems is the choice of an attribute [7] by which division will be made at a given node (partition attribute).…”

Section: Introductionmentioning

confidence: 99%

Two-level optimization by differential evolution in decision tree learning algorithm

Mitrofanov,

Semenkin

2024

ITM Web Conf.

View full text Add to dashboard Cite

Decision tree learning algorithms have a long history, but they are still popular due to their efficiency. Tree construction starts at the root and continues to each leaf node, creating a “near-optimal” tree. One of the key steps in creating a decision tree is the selection of a feature to split at each root node, which affects the classification accuracy. This process can be quite labor intensive. The article proposes a new approach to constructing decision trees based on the use of differential evolution for two-level tree optimization. Differential evolution works in two stages: at the first level, a feature is selected for separation, and at the second level, the threshold value for this feature is optimized. The results of the work were tested on several examples of classification problems.

show abstract

“…Further, consider the efficiency of differential evolution application in solving various problems of classification by decision trees. Two of the most popular decision tree training methods were chosen for the study, i.e., ID3 and CART (Mitrofanov & Semenkin, 2021). The author calls an algorithm where the selection of the threshold value is carried out by enumeration as a basic algorithm.…”

Section: Efficiency Of Differential Evolution Application In Solving ...mentioning

confidence: 99%

Hybridization of Machine Learning Models and Differential Evolution in Data Mining

Karaseva¹

2023

European Proceedings of Computers and Technology

View full text Add to dashboard Cite

Nowadays, data mining methods are applied in various fields of science and technologies. The widespread application of these methods necessitates the development of procedures to improve the efficiency of basic methods. Often, the efficiency of approaches based on data mining methods significantly depends on the selection of numerical coefficients. Moreover, these numerical coefficients can be both parameters of the algorithm and parameters of the obtained solutions. Therefore, it is possible to apply optimization methods to determine the best values. Nowadays, one of the most efficient methods of real optimization is differential evolution. The paper presents the study the differential evolution application to optimize coefficients of a dynamic system model obtained by an approach based on a self-configuring genetic programming algorithm. Also, it considers optimization of threshold values for solving classification problems with decision trees. The paper considers the most popular methods for training decision trees. The selected tasks demonstrate the universality of the proposed modification, since they solve diverse tasks.Moreover, the applied data analysis methods belong to different classes. The results of numerical experiments are presented for these tasks. They prove efficiency of the proposed hybridization.

show abstract

Tree retraining in the decision tree learning algorithm

Cited by 11 publications

References 3 publications

Hybrid Evolutionary Approach to Decision Trees Ensembles Design

Hybrid Evolutionary Approach to Decision Trees Ensembles Design

Two-level optimization by differential evolution in decision tree learning algorithm

Hybridization of Machine Learning Models and Differential Evolution in Data Mining

Contact Info

Product

Resources

About