How is your data arranged?

  • X,Y,Z,...,X,Y,Z,Label
  • X,X...,Y,Y...Z,Z...,Label
  • precomputed stats,Label
  • X,X,X...,Label (Only one axis)

What statistics should be computed?

Statistic All X Y Z
Mean
Std
Skewness
Kurtosis
Max
Min
Norm
Cov
r
DBA
ODBA
mean-diff
std-diff
wave amplitude
line crossings
25%
50%
75%

Select models to compute:

  • K Nearest Neighbors. choose k:
  • Linear SVM
  • RBF kernel SVM
  • Decision Tree
  • Random Forrest
  • Gaussian Naive Bayes
  • LDA
  • QDA
  • ANN
  • Select All

Feature Selection:

  • Use all features
  • F-test
  • PCA. Number of dimensions to use:

Select validation type:

  • Train-Test split
  • k-Fold Cross Validation. Choose k:
  • Leave One Out Cross Validation

You selected:

File name:
Data arranged as:
Models:
Feature Selection:
Validation: