Implementation of deep learning models in decoding fMRI data in a context of semantic processing
- To illustrate basic concepts of deep neural network models (DNN and CNN implementation so far)
- To illustrate how to utilize deep neural network models to decode fMRI data
- To illustrate principle ways of cross-validating the deep neural network models
- To illustrate how upsupervised learning model is more straightforward to implement in deep neural network models than scit-kit learning implementation
- To Propose a way to visualize the learned latent representations of the unsupervised models
- To illustrate simulation experiments with CNN models
- To illustrate representational similarity analysis
- To illustrate the implementation of searchlight algorithm
- To show how the literature has been using the deep neural network models in classification
- To demonstrate how deep neural network models can decode conscious states