| 程慧杰,陈 滨,刘芷余,何 颖,卜宪庚,高 越.肿瘤亚型识别研究中智能算法的应用[J].现代生物医学进展英文版,2019,19(5):960-964. |
| 肿瘤亚型识别研究中智能算法的应用 |
| Application of An Intelligent Algorithm in Tumor Subtype Recognition |
| Received:December 08, 2018 Revised:December 31, 2018 |
| DOI：10.13241/j.cnki.pmb.2019.05.037 |
| 中文关键词: 特征基因 BP神经网络 粒子群优化算法 肿瘤亚型识别 集成分类器 |
| 英文关键词: Feature gene BP neural network Particle swarm optimization (PSO) Tumor subtype recognition Ensemble classifier |
| 基金项目:黑龙江省教育厅科学技术研究项目（12521258） |
| Author Name | Affiliation | E-mail | | CHENG Hui-jie | Basic Medical College, Harbin Medical University, Harbin, Heilongjiang, 150086, China | 77050957@qq.com | | CHEN Bin | Basic Medical College, Harbin Medical University, Harbin, Heilongjiang, 150086, China | | | LIU Zhi-yu | Basic Medical College, Harbin Medical University, Harbin, Heilongjiang, 150086, China | | | HE Ying | Basic Medical College, Harbin Medical University, Harbin, Heilongjiang, 150086, China | | | BU Xian-geng | Basic Medical College, Harbin Medical University, Harbin, Heilongjiang, 150086, China | | | GAO Yue | The Fourth Affiliated Hospital of Harbin Medical University, Harbin, Heilongjiang, 150001, China | |
|
| Hits: 773 |
| Download times: 784 |
| 中文摘要: |
| 摘要 目的：为解决肿瘤亚型识别过程中易出现的维数灾难和过拟合问题，提出了一种改进的粒子群BP神经网络集成算法。方法：算法采用欧式距离和互信息来初步过滤冗余基因，之后用Relief算法进一步处理，得到候选特征基因集合。采用BP神经网络作为基分类器，将特征基因提取与分类器训练相结合，改进的粒子群对其权值和阈值进行全局搜索优化。结果：当隐含层神经元个数为5时，候选特征基因个数为110时，QPSO/BP算法全局优化和搜索，此时的分类准确率最高。结论：该算法不但提高了肿瘤分型识别的准确率，而且降低了学习的复杂度。 |
| 英文摘要: |
| ABSTRACT Objective: In order to solve the dimension disaster and over-fitting problems in the process of tumor subtype recogni- tion, a particle swarm optimization (PSO) BP neural network ensemble algorithm was proposed. Methods: The Euclidean distance and mutual information was used to preliminarily filter redundant genes, and then Relief algorithm was adopted to further process the candi- date feature genes set. The BP neural network was used as the base classifier, which combines feature genes extraction with classifier training. Results: When the number of hidden layer neurons is 5 and the number of candidate feature genes is 110, the QPSO/BP algo- rithm can optimize and search globally. Conclusion: The algorithm not only improves the accuracy of tumor classification and recogni- tion, but also reduces the complexity of learning. |
|
View Full Text
View/Add Comment Download reader |
| Close |
|
|
|
