Soft-split sparse regression based random forest for predicting future clinical scores of Alzheimer’s disease

Lei Huang, Yaozong Gao, Yan Jin, Kim Han Thung, Dinggang Shen

Research output: Chapter in Book/Report/Conference proceedingConference contribution

6 Citations (Scopus)

Abstract

In this study, we propose a novel sparse regression based random forest (RF) to predict future clinical scores of Alzheimer’s disease (AD) with the baseline scores and the MRI features. To avoid the stair-like decision boundary caused by axis-aligned split function in the conventional RF, we present a supervised method to construct the oblique split function by using sparse regression to select the informative features and transform the original features into the target-like features that are more discriminative. Then, we construct the oblique splitting function by applying the principal component analysis (PCA) on the transformed target-like features. Furthermore, to reduce the negative impact of potential missplit induced by the conventional “hard-split”, we further introduce the “soft-split” technique, in which both left and right nodes are visited with certain weights given a test sample. The experiment results show that sparse regression based RF alone can improve the prediction performance of the conventional RF. And further improvement can be achieved when both of the techniques are combined.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
PublisherSpringer Verlag
Pages246-254
Number of pages9
Volume9352
ISBN (Print)9783319248875
DOIs
Publication statusPublished - 2015
Externally publishedYes
Event6th International Workshop on Machine Learning in Medical Imaging, MLMI 2015 and Held in Conjunction with 18th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2015 - Munich, Germany
Duration: 2015 Oct 52015 Oct 5

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9352
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other6th International Workshop on Machine Learning in Medical Imaging, MLMI 2015 and Held in Conjunction with 18th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2015
CountryGermany
CityMunich
Period15/10/515/10/5

Fingerprint

Alzheimer's Disease
Random Forest
Regression
Oblique
Stairs
Magnetic resonance imaging
Principal component analysis
Target
Performance Prediction
Principal Component Analysis
Baseline
Transform
Predict
Vertex of a graph
Experiments
Experiment

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Huang, L., Gao, Y., Jin, Y., Thung, K. H., & Shen, D. (2015). Soft-split sparse regression based random forest for predicting future clinical scores of Alzheimer’s disease. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9352, pp. 246-254). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9352). Springer Verlag. https://doi.org/10.1007/978-3-319-24888-2_30

Soft-split sparse regression based random forest for predicting future clinical scores of Alzheimer’s disease. / Huang, Lei; Gao, Yaozong; Jin, Yan; Thung, Kim Han; Shen, Dinggang.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 9352 Springer Verlag, 2015. p. 246-254 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9352).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Huang, L, Gao, Y, Jin, Y, Thung, KH & Shen, D 2015, Soft-split sparse regression based random forest for predicting future clinical scores of Alzheimer’s disease. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 9352, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9352, Springer Verlag, pp. 246-254, 6th International Workshop on Machine Learning in Medical Imaging, MLMI 2015 and Held in Conjunction with 18th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2015, Munich, Germany, 15/10/5. https://doi.org/10.1007/978-3-319-24888-2_30
Huang L, Gao Y, Jin Y, Thung KH, Shen D. Soft-split sparse regression based random forest for predicting future clinical scores of Alzheimer’s disease. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 9352. Springer Verlag. 2015. p. 246-254. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-24888-2_30
Huang, Lei ; Gao, Yaozong ; Jin, Yan ; Thung, Kim Han ; Shen, Dinggang. / Soft-split sparse regression based random forest for predicting future clinical scores of Alzheimer’s disease. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 9352 Springer Verlag, 2015. pp. 246-254 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{5705e30fcfa6439b8a9a549092d42551,
title = "Soft-split sparse regression based random forest for predicting future clinical scores of Alzheimer’s disease",
abstract = "In this study, we propose a novel sparse regression based random forest (RF) to predict future clinical scores of Alzheimer’s disease (AD) with the baseline scores and the MRI features. To avoid the stair-like decision boundary caused by axis-aligned split function in the conventional RF, we present a supervised method to construct the oblique split function by using sparse regression to select the informative features and transform the original features into the target-like features that are more discriminative. Then, we construct the oblique splitting function by applying the principal component analysis (PCA) on the transformed target-like features. Furthermore, to reduce the negative impact of potential missplit induced by the conventional “hard-split”, we further introduce the “soft-split” technique, in which both left and right nodes are visited with certain weights given a test sample. The experiment results show that sparse regression based RF alone can improve the prediction performance of the conventional RF. And further improvement can be achieved when both of the techniques are combined.",
author = "Lei Huang and Yaozong Gao and Yan Jin and Thung, {Kim Han} and Dinggang Shen",
year = "2015",
doi = "10.1007/978-3-319-24888-2_30",
language = "English",
isbn = "9783319248875",
volume = "9352",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "246--254",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Soft-split sparse regression based random forest for predicting future clinical scores of Alzheimer’s disease

AU - Huang, Lei

AU - Gao, Yaozong

AU - Jin, Yan

AU - Thung, Kim Han

AU - Shen, Dinggang

PY - 2015

Y1 - 2015

N2 - In this study, we propose a novel sparse regression based random forest (RF) to predict future clinical scores of Alzheimer’s disease (AD) with the baseline scores and the MRI features. To avoid the stair-like decision boundary caused by axis-aligned split function in the conventional RF, we present a supervised method to construct the oblique split function by using sparse regression to select the informative features and transform the original features into the target-like features that are more discriminative. Then, we construct the oblique splitting function by applying the principal component analysis (PCA) on the transformed target-like features. Furthermore, to reduce the negative impact of potential missplit induced by the conventional “hard-split”, we further introduce the “soft-split” technique, in which both left and right nodes are visited with certain weights given a test sample. The experiment results show that sparse regression based RF alone can improve the prediction performance of the conventional RF. And further improvement can be achieved when both of the techniques are combined.

AB - In this study, we propose a novel sparse regression based random forest (RF) to predict future clinical scores of Alzheimer’s disease (AD) with the baseline scores and the MRI features. To avoid the stair-like decision boundary caused by axis-aligned split function in the conventional RF, we present a supervised method to construct the oblique split function by using sparse regression to select the informative features and transform the original features into the target-like features that are more discriminative. Then, we construct the oblique splitting function by applying the principal component analysis (PCA) on the transformed target-like features. Furthermore, to reduce the negative impact of potential missplit induced by the conventional “hard-split”, we further introduce the “soft-split” technique, in which both left and right nodes are visited with certain weights given a test sample. The experiment results show that sparse regression based RF alone can improve the prediction performance of the conventional RF. And further improvement can be achieved when both of the techniques are combined.

UR - http://www.scopus.com/inward/record.url?scp=84951971797&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84951971797&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-24888-2_30

DO - 10.1007/978-3-319-24888-2_30

M3 - Conference contribution

AN - SCOPUS:84951971797

SN - 9783319248875

VL - 9352

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 246

EP - 254

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

PB - Springer Verlag

ER -