Multiple Imputation for Biomedical Data using Monte Carlo Dropout Autoencoders

被引:3
|
作者
Miok, Kristian [1 ]
Dong Nguyen-Doan [1 ]
Robnik-Sikonja, Marko [2 ]
Zaharie, Daniela [1 ]
机构
[1] West Univ Timisoara, Fac Math & Comp Sci, Dept Comp Sci, Timisoara, Romania
[2] Univ Ljubljana, Fac Comp & Informat Sci, Ljubljana, Slovenia
基金
欧盟地平线“2020”;
关键词
data preprocessing; missing data imputation; deep learning models; Monte Carlo dropout;
D O I
10.1109/ehb47216.2019.8969940
中图分类号
R19 [保健组织与事业(卫生事业管理)];
学科分类号
摘要
Due to complex experimental settings, missing values are common in biomedical data. To handle this issue, many methods have been proposed, from ignoring incomplete instances to various data imputation approaches. With the recent rise of deep neural networks, the field of missing data imputation has oriented towards modelling of the data distribution. This paper presents an approach based on Monte Carlo dropout within (Variational) Autoencoders which offers not only very good adaptation to the distribution of the data but also allows generation of new data, adapted to each specific instance. The evaluation shows that the imputation error and predictive similarity can be improved with the proposed approach.
引用
收藏
页数:4
相关论文
共 50 条
  • [1] Generating Data using Monte Carlo Dropout
    Miok, Kristian
    Dong Nguyen-Doan
    Zaharie, Daniela
    Robnik-Sikonja, Marko
    [J]. 2019 IEEE 15TH INTERNATIONAL CONFERENCE ON INTELLIGENT COMPUTER COMMUNICATION AND PROCESSING (ICCP 2019), 2019, : 509 - 515
  • [2] Leveraging Variational Autoencoders for Multiple Data Imputation
    Roskams-Hieter, Breeshey
    Wells, Jude
    Wade, Sara
    [J]. MACHINE LEARNING AND KNOWLEDGE DISCOVERY IN DATABASES: RESEARCH TRACK, ECML PKDD 2023, PT I, 2023, 14169 : 491 - 506
  • [3] Markov chain Monte Carlo multiple imputation using Bayesian networks for incomplete intelligent transportation systems data
    Ni, Daiheng
    Leonard, John D., II
    [J]. INFORMATION SYSTEMS AND TECHNOLOGY, 2005, (1935): : 57 - 67
  • [4] Multiple imputation of missing at random data: General points and presentation of a Monte-Carlo method
    Cottrell, G.
    Cot, M.
    Mary, J. -Y.
    [J]. REVUE D EPIDEMIOLOGIE ET DE SANTE PUBLIQUE, 2009, 57 (05): : 361 - 372
  • [5] Partial Multiple Imputation With Variational Autoencoders: Tackling Not at Randomness in Healthcare Data
    Pereira, Ricardo Cardoso
    Abreu, Pedro Henriques
    Rodrigues, Pedro Pereira
    [J]. IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS, 2022, 26 (08) : 4218 - 4227
  • [6] Physiological Waveform Imputation of Missing Data using Convolutional Autoencoders
    Miller, Daniel
    Ward, Andrew
    Bambos, Nicholas
    Scheinker, David
    Shin, Andrew
    [J]. 2018 IEEE 20TH INTERNATIONAL CONFERENCE ON E-HEALTH NETWORKING, APPLICATIONS AND SERVICES (HEALTHCOM), 2018,
  • [7] Imputation of Missing Traffic Flow Data Using Denoising Autoencoders
    Jiang, Boyuan
    Siddiqi, Muhammad Danial
    Asadi, Reza
    Regan, Amelia
    [J]. 12TH INTERNATIONAL CONFERENCE ON AMBIENT SYSTEMS, NETWORKS AND TECHNOLOGIES (ANT) / THE 4TH INTERNATIONAL CONFERENCE ON EMERGING DATA AND INDUSTRY 4.0 (EDI40) / AFFILIATED WORKSHOPS, 2021, 184 : 84 - 91
  • [8] A Markov chain Monte Carlo algorithm for multiple imputation in large surveys
    Daniel Schunk
    [J]. AStA Advances in Statistical Analysis, 2008, 92 : 101 - 114
  • [9] A Markov chain Monte Carlo algorithm for multiple imputation in large surveys
    Schunk, Daniel
    [J]. ASTA-ADVANCES IN STATISTICAL ANALYSIS, 2008, 92 (01) : 101 - 114
  • [10] Explainable Fingerprint ROI Segmentation Using Monte Carlo Dropout
    Joshi, Indu
    Kothari, Riya
    Utkarsh, Ayush
    Kurmi, Vinod K.
    Dantcheva, Antitza
    Roy, Sumantra Dutta
    Kalra, Prem Kumar
    [J]. 2021 IEEE WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION WORKSHOPS (WACVW 2021), 2021, : 60 - 69