matlab白化代码-csp_evaluation:评估不具有完整等级的协方差矩阵的常见空间模式实现

  • u6_389437
    了解作者
  • 147.3MB
    文件大小
  • zip
    文件格式
  • 0
    收藏次数
  • VIP专享
    资源类型
  • 0
    下载次数
  • 2022-05-14 10:43
    上传日期
matlab白化代码评估不具有完整等级的协方差矩阵的常见空间模式实现 内容 implementations :CSP实施 我们的实现: csp_python.py :所有Python实现 csp_geometric_approach_dim_reduction.m :在增白步骤中具有csp_geometric_approach_dim_reduction.m维功能的Matlab几何方法 csp_geometric_approach_no_checks.m :不降低维数的Matlab几何方法(不检查) csp_gep_no_checks.m :Matlab广义特征值问题处理方法(无检查) 工具箱: use_bbci.m :BBCI工具箱 use_fieldtrip.m :FieldTrip use_biosig.m :BioSig use_mne.py artifacts_removal :基于ICA的工件去除方法 evaluation.py :评估二进制运动图像分类任务中的所有CSP实现 复写 从environment.yml文件创建Conda环境: conda env create
csp_evaluation-master.zip
内容介绍
# Evaluation of common spatial patterns implementations for covariance matrices without full rank ### Content - `implementations`: CSP implementations - Our implementations: - `csp_python.py`: all Python implementations - `csp_geometric_approach_dim_reduction.m`: Matlab geometric approach with dimensionality reduction during the whitening step - `csp_geometric_approach_no_checks.m`: Matlab geometric approach without dimensionality reduction (no checks) - `csp_gep_no_checks.m`: Matlab generalized eigenvalue problem approach (no checks) - Toolboxes: - `use_bbci.m`: BBCI Toolbox - `use_fieldtrip.m`: FieldTrip - `use_biosig.m`: BioSig - `use_mne.py`: MNE - `artifacts_removal`: artifact removal methods based on ICA - `evaluation.py`: Evaluate all CSP implementations on a binary motor imagery classification task ## Replication Create the Conda environment from the `environment.yml` file: ```console conda env create -f environment.yml ``` Activate the new environment: ```console conda activate csp_evaluation ``` Note that CSP implementations from Matlab Toolboxes for EEG analysis contain hard-coded paths for toolboxes locations in their corresponding .m files. Directory `results` already contains pre-computed ICA for each patient. Evaluation can be run in parallel (on the same or multiple computers), which is useful for Matlab implementations, by uncommenting line 166 in `evaluation.py` ```python # @TaskGenerator ``` Then, tasks will be executed using [jug](https://jug.readthedocs.io/en/latest/) package. Run it by ```console jug execute evaluation.py ``` and see the progress by ```console jug status evaluation.py ``` Results can be inspected and visualized by scripts in `visualize` directory.
评论
    相关推荐