ACC
Feature Diagnosis of Alzheimer’s Disease Based on Stacking Fusion Algorithm and DBSCAN Clustering
TOTAL VIEWS: 2932
Abstract
Alzheimer's disease (AD) is a progressive neurodegenerative disease. It is characterized by a progressive decline in the ability to perform activities of daily living, accompanied by various neuropsychiatric symptoms and behavioral disorders. The disease usually develops progressively in older people, with progressive loss of independent living and death from complications 10 to 20 years after onset. First of all, we eliminate the indicators that the data vacancy accounts for 80% or more of the data volume, and eliminate similar duplicate indicators. Collaborative filtering algorithm and LOF abnormal data detection algorithm are used to process the data and get the original data. The importance of establishing a random forest model to calculate characteristic indexes, and combining Spearman correlation coefficient method to solve the correlation between the characteristics of original data and the diagnosis of Alzheimer's disease. Then divide the data into five categories. XGboost and SVM models with 50% cross-validation in stack fusion are used to classify characteristic indexes, study cognitive behavior and structural characteristics of brain, and design intelligent diagnosis of Alzheimer's disease. And we divided CN, MCI and AD into three categories, and screened the index data in nine categories. Unsupervised classification is carried out by DBSCAN clustering, and three subclasses are further refined into three subclasses by elbow method. Then, according to the classification results, the characteristic indexes and patients were analyzed by single factor, and some indexes were eliminated. The change trend of characteristic indicators with time was studied, and the evolution patterns of different types of diseases with time were obtained by combining descriptive statistics of indicators.
Keywords
Random forest algorithm, XGboost, stacking Fusion algorithm, DBSCAN, Elbow method
References
[1] I Shoukui, Mathematical Modeling Algorithm and Program [M]. Naval Aeronautical Engineering Institute, 2007.
[2] Huo Jinwu, Application of MATLAB in Mathematical Modeling [M]. Beijing University of Aeronautics and Astronautics Press, 2014.
[3] Han Zhonggeng, Mathematical Modeling Method and Its Application [M]. Beijing: Higher Education Press, 2017.
[4] I Xin, MATLAB Mathematical Modeling [M]. Beijing: Tsinghua University Press, 2017.
[5] Hen Lichang, Heng Mingli, Wang Jun, Chen Pingyan. A comparative study of quantitative longitudinal missing data processing methods based on multiple missing mechanisms [J]. Modern Preventive Medicine, 2020, 47(20):3684-3687+3697.
[6] Ang Guang, Wang Lei, Fu Zhihui. Influence of Different missing Data Processing Methods on D-vine Copula Classifier [J]. Journal of Shenyang Normal University (Natural Science Edition), 2021, 39(01):35-38.
[7] Wang Zhaoying. Research on the performance of listed companies on the New Third Board based on multivariate statistical analysis [J]. Popular Standardization, 2021(15):109-111.
[8] Hang Shuai. Research on the correlation between B2B E-commerce transaction and economic growth [J]. Journal of Changchun University, 2012, 22(07):806-808.
[9] Huang Ying, Zhang Junxi. Research on E-commerce Sales Forecasting Based on GM(1,N)-Prophet Model [J]. Journal of Southwest University for Nationalities (Natural Science Edition), 2021, 47(03):317-325.
[10] Lin Jiaming. Research on Statistical Indicators and Accounting Methods of E-commerce [D]. South China University of Technology, 2015.
[11] LUO R. Research on Sales Prediction Model of chain Enterprise based on machine learning [D]. Chongqing University of Science and Technology, 2022. DOI: 10.27753 /, dcnki. GCQGX. 2022.000227.
[12] SUN Mengjing. Discussion on the application of XBRL financial Report in Chinese listed commercial banks [D]. Jiangxi University of Finance and Economics, 2014.
[13] Wu Zhongsheng, Zhang Tianxi, Li Zhengzheng. Research on XBRL technology from the perspective of business report supply chain [J]. Modern Management Science, 2013(07):91-93.
How to cite this paper
Feature Diagnosis of Alzheimer’s Disease Based on Stacking Fusion Algorithm and DBSCAN Clustering
How to cite this paper: Wenchen Wu, Bin Ma, Jingyi Zhu, Jiajun Huang. (2023) Feature Diagnosis of Alzheimer’s Disease Based on Stacking Fusion Algorithm and DBSCAN Clustering. Advances in Computer and Communication, 4(3), 143-148.
DOI: http://dx.doi.org/10.26855/acc.2023.06.007
More Articles
- Research on Equipment Support Command Issues Based on Intelligent Decision-making
- Construction and Analysis of Disease Networks Based on Multi-omics Data
- Research on Causal Reasoning Method for MOOC Course Effectiveness Based on Multi-path Network Search
- Study on Integrated Watershed Management Decision-making Based on Artificial Intelligence
- AI Application in the Logistics Industry
- Research on the Application of Intelligent Tourism Data Analysis Based on Python