Back

HiddenVis: a Hidden State Visualization Toolkit to Visualize and Interpret Deep Learning Models for Time Series Data

Yan, J.; Rong, R.; Xiao, G.; Zhan, X.

2020-12-12 bioinformatics
10.1101/2020.12.11.422030 bioRxiv
Show abstract

Accelerometers provide continuous measurements of human movements. Utilizing their data exhibits great potential for monitoring or predicting human health outcomes. In this direction, researchers are active in developing machine-learning algorithms. However, compared to understanding deep learning models, visualizing and interpreting such models is equally important but has been less well-developed. To address these limitations, we constructed an online software tool to visualize the hidden states inferred from the deep learning models and to compare different cohorts based on outcome-associated temporal profiles. We demonstrated the utility of the model using the National Health and Nutrition Examination Survey (NHANES) dataset. Using this tool, we discovered that two time periods of daily activity contribute to significant differences in 5-year mortalities. To disseminate the software to the broader community for the analysis of accelerometer data, we provide the work as open-source code at https://github.com/jyan97/HiddenVis.

Matching journals

The top 8 journals account for 50% of the predicted probability mass.

1
PLOS ONE
● 5266 papers in training set
Top 13%
15.0%
Similar papers in this journal
2
BMC Bioinformatics
○ 457 papers in training set
Top 1%
6.7%
Similar papers in this journal
3
JMIR mHealth and uHealth
◐ 11 papers in training set
Top 0.1%
6.2%
Similar papers in this journal
4
PLOS Computational Biology
● 1863 papers in training set
Top 7%
5.5%
Similar papers in this journal
5
Journal of Medical Internet Research
◐ 87 papers in training set
Top 0.4%
5.1%
Similar papers in this journal
6
IEEE Access
● 35 papers in training set
Top 0.2%
4.8%
7
GigaScience
◐ 212 papers in training set
Top 0.8%
4.3%
8
Bioinformatics
◐ 1204 papers in training set
Top 5%
4.0%
50% of probability mass above
9
Journal of the American Medical Informatics Association
◐ 71 papers in training set
Top 0.8%
4.0%
10
IEEE Journal of Biomedical and Health Informatics
● 37 papers in training set
Top 0.3%
3.4%
11
BMC Medical Informatics and Decision Making
○ 43 papers in training set
Top 0.7%
2.8%
12
PLOS Digital Health
● 106 papers in training set
Top 2%
2.4%
13
Scientific Reports
○ 3612 papers in training set
Top 42%
2.4%
14
Computers in Biology and Medicine
○ 128 papers in training set
Top 2%
2.1%
15
JAMIA Open
◐ 42 papers in training set
Top 0.9%
1.7%
16
Journal of Biomedical Informatics
○ 47 papers in training set
Top 0.8%
1.7%
17
Patterns
○ 78 papers in training set
Top 1%
1.7%
18
DIGITAL HEALTH
○ 17 papers in training set
Top 0.5%
1.3%
19
PeerJ
◐ 308 papers in training set
Top 7%
1.3%
20
Sensors
○ 43 papers in training set
Top 0.9%
1.1%
21
eLife
● 5828 papers in training set
Top 59%
1.1%
22
Bioinformatics Advances
◐ 203 papers in training set
Top 4%
1.1%
23
Nature Communications
○ 5641 papers in training set
Top 52%
1.1%
24
Expert Systems with Applications
○ 11 papers in training set
Top 0.3%
1.0%
25
International Journal of Medical Informatics
○ 26 papers in training set
Top 1%
1.0%
26
Database
◐ 61 papers in training set
Top 0.7%
1.0%
27
Biology Methods and Protocols
◐ 61 papers in training set
Top 2%
0.8%
28
Computational and Structural Biotechnology Journal
● 242 papers in training set
Top 7%
0.8%
29
Cerebral Cortex
◐ 396 papers in training set
Top 5%
0.6%
30
Current Protocols
○ 15 papers in training set
Top 0.3%
0.6%

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.