search

cnhfeng / RASNet

"RASNet: Recurrent Aggregation Neural Network for Safe and Efficient Drug Recommendation"
4 stars 0 forks source link

About Case study #1

Open FansonWang opened 3 months ago

FansonWang commented 3 months ago

Thank you very much for doing such a excellent job.Your case study experiment was done very well. Could you please tell me how you implement it in detail？Thank you!

cnhfeng commented 3 months ago

Thank you for supporting our work. For Case Study, we randomly selected a patient with 5 or more visits in the test set as a test sample, and then loaded the parameters of the other models (if their trained .model files are not given in the model's code, here you need to train the model and select the .model file with the best results). After the test sample is entered into the model, we get the probability values recommended by each model, we use the method of finding the DDI rate in the model to find the DDI rate for each drug combination, and we then use the med_voc dictionary to restore its ATC4 identifier. We get the number of correct ones and the number of incorrect ones by comparing them with the real labels. Thank you for your question.