AUTHOR=Ji Naihua , Bao Rongyi , Mu Xiaoyi , Chen Zhao , Yang Xin , Wang Shumei 

TITLE=RETRACTED: Cost-sensitive classification algorithm combining the Bayesian algorithm and quantum decision tree

JOURNAL=Frontiers in Physics

VOLUME=Volume 11 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/physics/articles/10.3389/fphy.2023.1179868

DOI=10.3389/fphy.2023.1179868

ISSN=2296-424X

ABSTRACT=The fundamental issue of quantum data mining classification, the quantum classifier, has emerged as a key topic in machine learning.  However, current quantum classifiers still have some drawbacks such as the inability to process incremental data and failure to take the cost of categorization into account.  These flaws have developed into critical issues that need to be resolved by quantum classification methods.  The majority of data classification techniques currently in use are classical machine learning or deep learning algorithms, but these techniques are inefficient and have limited data processing capabilities. The ability of categorization algorithms to handle incremental data in the big data environment has also become a priority since the emergence of the big data era.  In this study, we propose a global decision tree paradigm to address these issues.  We aim to design a complete quantum decision tree classification algorithm that has high accuracy and efficiency, while maintaining high performance when processing incremental sequences and considering classification costs.  Based on data objects and cost constraints, the model generates a suitable decision tree dynamically. To handle incremental data, we integrate the Bayesian algorithm and the quantum decision tree classification algorithm in this study.  By adding kernel functions obtained from quantum kernel estimation to a linear quantum support vector machine (QSVM), we construct a decision tree classifier employing decision-directed acyclic networks of QSVM nodes (QKE).  Our studies show that, in terms of classification accuracy, speed, and practical application impact, our classification approach outperforms the competition. The effectiveness and adaptability of our algorithm are confirmed, with its accuracy difference from conventional classification algorithms being less than 1\%.  With improved accuracy and reduced expense as the incremental data increases, the efficiency of our suggested algorithm for incremental data classification is comparable to previous quantum classification algorithms.  Theoretical and experimental findings demonstrate the effectiveness of the suggested quantum classification technique.