来源：["合肥高新杯"心电人机智能大赛](https://tianchi.aliyun.com/competition/entrance/231754/introduction)
**A版单模型20190916在线F1-Score=0.801**
**B版单模型20191010在线F1-Score=0.837**
系统环境：*centos7 python3.6 pytorch1.0*
**大致思路：每个导联作为一个通道，使用1维卷积进行训练**
# 数据预处理
数据解压放在data目录下，使用8个导联的数据，简单进行train_val数据集划分
```shell
python data_process.py
```
# 模型训练
```shell
python main.py train #从零开始训练
```
# 模型测试
模型测试，在submit文件夹下生成提交结果
```shell
python main.py test --ckpt=..model_path #加载预训练权重进行测试
```
**一些细节**
1. 本次测试模型为1dconv_resnet34，直接修改于torchvision
2. 训练数据只进行了简单的数据增强，最终无normalize
3. 由于设备问题，数据进行了重采样，推荐使用原数据
参考论文：
https://www.nature.com/articles/s41591-018-0268-3
心电智能比赛.zip （86个子文件） 
cm 
utils.py 1KB .vscode launch.json 494B label_and_example submit_example_1217.csv 885KB train_label_1217.csv 2.02MB dataset_split.ipynb 3KB train.csv 2.28MB topredict_id.csv 442KB .ipynb_checkpoints train_label_1217-checkpoint.csv 2.02MB dataset_split-checkpoint.ipynb 72B test-checkpoint.csv 773KB train-checkpoint.csv 2.28MB submit_example_1217-checkpoint.csv 885KB test.csv 773KB main.py 8KB LICENSE 11KB data_split_directory.py 988B submit subA_202112250013.csv 0B subA_202112252356.csv 873KB subA_202112251327.csv 873KB subA_202112241729.csv 873KB subA_202112271046.csv 873KB subA_202112242129.csv 873KB subA_integrate_202112262348.csv 0B subA_202112251147.csv 0B subA_202112251313.csv 0B subA_integrate_202112262149.csv 0B subA_202112260910.csv 873KB subA_202112250035.csv 873KB subA_202112262348.csv 759B subA_integrate_202112271028.csv 0B subA_integrate_202112262151.csv 0B subA_integrate_202112262344.csv 0B subA_202112260008.csv 873KB subA_202112270945.csv 0B subA_202112251143.csv 0B subA_integrate_202112271050.csv 0B subA_202112260000.csv 0B subA_202112260750.csv 0B subA_202112251320.csv 0B subA_202112242107.csv 0B subA_202112252345.csv 0B subA_202112270957.csv 873KB subA_integrate_202112262347.csv 0B .ipynb_checkpoints subA_202112270957-checkpoint.csv 873KB subA_integrate_202112262151-checkpoint.csv 0B subA_202112271046-checkpoint.csv 873KB subA_integrate_202112271028-checkpoint.csv 0B subA_202112270945-checkpoint.csv 0B subA_202112271108-checkpoint.csv 873KB subA_integrate_202112262345.csv 0B subA_202112271108.csv 873KB dataset.py 4KB example.py 148B models __init__.py 175B rnn.py 1KB resnet.py 7KB SiT.py 5KB __pycache__ resnet.cpython-38.pyc 6KB rnn.cpython-38.pyc 1KB __init__.cpython-38.pyc 363B SiT.cpython-38.pyc 5KB .ipynb_checkpoints resnet-checkpoint.py 7KB SiT-checkpoint.py 4KB rnn-checkpoint.py 1KB __init__-checkpoint.py 175B .gitignore 20B run.sh 62B __pycache__ test.cpython-38.pyc 1KB data_process.cpython-38.pyc 3KB config.cpython-38.pyc 941B utils.cpython-38.pyc 2KB dataset.cpython-38.pyc 4KB demo.py 261B test.py 9KB .ipynb_checkpoints main-checkpoint.py 8KB test-checkpoint.py 9KB data_process-checkpoint.py 2KB config-checkpoint.py 1KB utils-checkpoint.py 1KB data_split_directory-checkpoint.py 988B example-checkpoint.py 349B dataset-checkpoint.py 4KB README.md 1KB launch.json 495B data_process.py 2KB config.py 1KB资源评论
