README

A dataset level implementation of GRAD-CAM++ for time series and 1d signal

Installation

!pip install dataset_gradcam_plus_plus
                 or
 cd ./dataset_gradcam_plus_plus  python setup.py install

Local classification explaination

from dataset_gradcam_plus_plus import local_features_importances

local_features_importances(model,data,time,layer_name,label)
Arguments:
- model : The model trained
- data : The data sample
- time : The time range. The user can create a linspace vector and use it as the time range.
- layer_name : The last layer of the feature extraction part of the model. Usually, it is the last layer before the Flattening operation.
- label : The label output list

![Alt text](https://github.com/marcusnk237/dataset_gradcam_plus_plus/blob/main/results/gradcam_plus_plus_1d.png)

Dataset level feature relevance

from dataset_gradcam_plus_plus import dataset_features_importances
def dataset_features_importances(model,datas,layer_name,feature_names,flag=True)

Arguments:
- model : The model trained
- datas : the dataset. Due of Hardware limitations, the max size of datas is 5000 samples.
- layer_name : The last layer of the feature extraction part of the model. Usually, it is the last layer before the Flattening operation.
- feature_names : The feature list
- flag : True, return the global feature importance plot; False, return the feature importance frequency and the corresponding features

![Alt text](https://github.com/marcusnk237/dataset_gradcam_plus_plus/blob/main/results/dataset_level_feature_importance.jpg)

Sources : 
The original Grad-CAM ++ publication :https://arxiv.org/pdf/1710.11063.pdf

License
This project is Licensed under the MIT License.