I am working on EEG classification problem and have collected my own data set ,at first step I applied filters and ICA than decomposed it by applying db3 at level 6 ,after trying all kind of known feature(Non Statistical i.e. Hjorth features, Entropy, band Power and Statistical features mean, median, variance ,skewness ,std_deviation ) accuracy with SVM classifier with 10-K fold method has stuck at 63% ,where as when i fed the chunk of cleaned and pruned data to SVM the accuracy of different chunks is 80% .what should i report in paper please guide.
In your paper, u should comprehensively report both the traditional feature extraction approach and ur novel method of directly using cleaned and pruned EEG data with the SVM classifier. Here's what to include:
This structure will help underline the efficacy and reliability of your approach providing valuable insights into EEG data classification methodologies...
- Badges
- Science topic
- Similar topics
- Methodology
- Research Methodology
- Research Papers