Denoising Autoencoder and Weight Initialization of CNN Model for ERP Classification

Keyphrases

• Event-related Potentials 100%
• Convolutional Neural Network Model 100%
• Denoising Autoencoder 100%
• Weight Initialization 100%
• Brain-computer Interface Systems 50%
• Electroencephalogram Signals 50%
• Deep Learning 25%
• Superior Performance 25%
• Deep Neural Network 25%
• Temporal Domain 25%
• Brain-computer Interface 25%
• Spatial Domain 25%
• Signal-to-noise Ratio 25%
• Linear Classifier 25%
• Denoising 25%
• Feature Space 25%
• Potential-based 25%
• Parameter Dependence 25%
• Learning Process 25%
• Initialize 25%
• Optimal Subset 25%
• Laplacian 25%
• Decoding Accuracy 25%
• Spelling Systems 25%
• Unsupervised Learning 25%
• Event-related Potential Signal 25%
• Spelling Performance 25%
• Autoencoder 25%
• Target Character 25%
• Single-trial Decoding 25%
• End-to-end Deep Learning Architecture 25%
• Popular 25%
• Filtering Algorithm 25%
• Preprocessing Step 25%

INIS

• classification 100%
• weight 100%
• neural networks 100%
• signals 60%
• computers 60%
• interfaces 60%
• brain 60%
• performance 40%
• accuracy 40%
• algorithms 40%
• learning 40%
• people 20%
• implementation 20%
• space 20%
• risks 20%
• architecture 20%
• pca 20%
• laplacian 20%
• signal-to-noise ratio 20%

Computer Science

• Neural Network Model 100%
• Convolutional Neural Network 100%
• Denoising Autoencoder 100%
• Computer Interface 60%
• Interface System 40%
• Electroencephalogram 40%
• Superior Performance 20%
• Deep Neural Network 20%
• Unsupervised Learning 20%
• de-noising 20%
• Feature Space 20%
• Learning Process 20%
• Linear Classifier 20%
• Laplace Operator 20%
• Preprocessing Step 20%
• Spelling Performance 20%
• Spatial Domain 20%
• Deep Learning Method 20%
• Autoencoder 20%
• Signal-to-Noise Ratio 20%