Datasets:
Tasks:
Tabular Classification
Modalities:
Tabular
Formats:
csv
Sub-tasks:
tabular-multi-class-classification
Size:
100K - 1M
Tags:
healthcare
diabetes
readmission
electronic-health-records
uncertainty-quantification
clinical-prediction
License:
Youran Li commited on
Upload DPR dataset splits and data card
Browse files- .gitattributes +2 -0
- README.md +49 -0
- cohort.csv +0 -0
- features.csv +3 -0
- test.csv +0 -0
- train.csv +3 -0
- validation.csv +0 -0
.gitattributes
CHANGED
|
@@ -58,3 +58,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 58 |
# Video files - compressed
|
| 59 |
*.mp4 filter=lfs diff=lfs merge=lfs -text
|
| 60 |
*.webm filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
| 58 |
# Video files - compressed
|
| 59 |
*.mp4 filter=lfs diff=lfs merge=lfs -text
|
| 60 |
*.webm filter=lfs diff=lfs merge=lfs -text
|
| 61 |
+
features.csv filter=lfs diff=lfs merge=lfs -text
|
| 62 |
+
train.csv filter=lfs diff=lfs merge=lfs -text
|
README.md
ADDED
|
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: unknown
|
| 3 |
+
task_categories:
|
| 4 |
+
- tabular-classification
|
| 5 |
+
tags:
|
| 6 |
+
- healthcare
|
| 7 |
+
- tabular
|
| 8 |
+
- readmission
|
| 9 |
+
pretty_name: Diabetic 30-Day Readmission Dataset
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# Dataset Card for DPR
|
| 13 |
+
|
| 14 |
+
This dataset is derived from the diabetic readmission dataset and is formatted for binary prediction of 30-day hospital readmission.
|
| 15 |
+
|
| 16 |
+
## Prediction Task
|
| 17 |
+
|
| 18 |
+
The target label is `readmit_30d`.
|
| 19 |
+
|
| 20 |
+
- `1`: patient was readmitted within 30 days (`readmitted == "<30"`)
|
| 21 |
+
- `0`: patient was not readmitted within 30 days
|
| 22 |
+
|
| 23 |
+
## Files
|
| 24 |
+
|
| 25 |
+
- `train.csv`: training split
|
| 26 |
+
- `validation.csv`: validation split
|
| 27 |
+
- `test.csv`: test split
|
| 28 |
+
- `cohort.csv`: patient/encounter identifiers and labels
|
| 29 |
+
- `features.csv`: feature columns excluding the target label
|
| 30 |
+
|
| 31 |
+
## Splits
|
| 32 |
+
|
| 33 |
+
| Split | N | Positives | Positive Rate |
|
| 34 |
+
|---|---:|---:|---:|
|
| 35 |
+
| Train | 71236 | 7950 | 0.1116 |
|
| 36 |
+
| Validation | 15265 | 1704 | 0.1116 |
|
| 37 |
+
| Test | 15265 | 1703 | 0.1116 |
|
| 38 |
+
|
| 39 |
+
## Preprocessing
|
| 40 |
+
|
| 41 |
+
Missing values originally encoded as `?` were replaced with missing values. The binary label `readmit_30d` was created from the original `readmitted` column. The data were split into 70/15/15 train/validation/test partitions using stratification on `readmit_30d`.
|
| 42 |
+
|
| 43 |
+
## Intended Use
|
| 44 |
+
|
| 45 |
+
This dataset is intended for evaluating binary classification and uncertainty quantification methods for 30-day readmission prediction.
|
| 46 |
+
|
| 47 |
+
## Limitations
|
| 48 |
+
|
| 49 |
+
This dataset should not be used for clinical decision-making without further validation. Potential limitations include retrospective design, missingness, class imbalance, and dataset-specific biases.
|
cohort.csv
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
features.csv
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:95066ffc9ab86187c77c605e8c9907b9fb908932d5394159bba28c5412959e9c
|
| 3 |
+
size 17787884
|
test.csv
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
train.csv
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4af2d4be6198c5529e0676b4ce9450c42bc5e29efe6ff6c43956c146089ef151
|
| 3 |
+
size 12839394
|
validation.csv
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|