I have to build a hybrid model for prediction of heart disease. Before that, i want everyone of you to clarify me few things.
[1] Is it possible to pass output of 1 classifier to other classifier? like SVM-> ANN -> RF -> LR
- According to me, it is possible but label/target data will be reduce when passed from 1 classifier to other classifier. In other classifier, we have to give same size of training data, so that size of data will be matched. But that a trick.
- suggest me some paper/code, if you have anything like that.
[2] What are parameter should i consider while building Hybrid model?
- i am aware of Kappa value but suggest me more
[3] On what basis should i select algorithms for Hybrid Model?
[4] What are right practices or common methods for building Hybrid Model?
- i know ensemble learning but suggestion are welcome.
any hybrid models, such as ANN-based, ANFIS-based, ELM-based, SVM-based, RVM-based, LSSVM-based, and so on.
Suggest you to go for deep learning models, such as LSTM< RNN, Bi-LSTM, etc.
- Badges
- Science topic
- Similar topics
- Organizational Psychology
- Training