Skip to main content
Cross Validated

Return to Question

Tag additionand minor grammaticall fix.
Source Link
usεr11852
usεr11852
  • 46k
  • 3
  • 106
  • 165

I have problems understanding data augmentation. I currently have low dimension-dimension features, each data point only has 3 features. A total of 20k non-linear data with only 3 features. I have generated about 120 additional features from statistical values (e.g. mean, var, std). Is generating statistical values called data augmentation? Is it possible for the augmented data to be oversampled like SMOTE? What would be the validity?

I have problems understanding data augmentation. I currently have low dimension features, each data point only has 3 features. A total of 20k non-linear data with only 3 features. I have generated about 120 additional features from statistical values (e.g. mean, var, std). Is generating statistical values called data augmentation? Is it possible for the augmented data to be oversampled like SMOTE? What would be the validity?

I have problems understanding data augmentation. I currently have low-dimension features, each data point only has 3 features. A total of 20k non-linear data with only 3 features. I have generated about 120 additional features from statistical values (e.g. mean, var, std). Is generating statistical values called data augmentation? Is it possible for the augmented data to be oversampled like SMOTE? What would be the validity?

Improved wording
Source Link
edit approved Oct 25, 2022 at 13:30
Lynn
edit approved Oct 25, 2022 at 13:30
Lynn
  • 1.8k
  • 1
  • 4
  • 20

I have problemproblems understanding data augmentation. I am currently I have low dimension features, each data point only havehas 3 features. A total of 20k non-linear data with only have 3 features. I have generated about 120 additional features from statistical valuevalues (meane.g. mean, var, std). DoesIs generating statistical valuevalues called data augmentation? Is it possible for the augmented data to be oversampleoversampled like SMOTE? What would be the validity?

I have problem understanding data augmentation. I am currently I have low dimension features, each data only have 3 features. A total of 20k non-linear data only have 3 features. I have generated about 120 additional features from statistical value (mean, var, std). Does generating statistical value called data augmentation? Is it possible the augmented data to be oversample like SMOTE? What would be the validity?

I have problems understanding data augmentation. I currently have low dimension features, each data point only has 3 features. A total of 20k non-linear data with only 3 features. I have generated about 120 additional features from statistical values (e.g. mean, var, std). Is generating statistical values called data augmentation? Is it possible for the augmented data to be oversampled like SMOTE? What would be the validity?

Source Link
kasyful
kasyful
  • 23
  • 5

Synthetic vs Data augmentation for low dimensionality data

I have problem understanding data augmentation. I am currently I have low dimension features, each data only have 3 features. A total of 20k non-linear data only have 3 features. I have generated about 120 additional features from statistical value (mean, var, std). Does generating statistical value called data augmentation? Is it possible the augmented data to be oversample like SMOTE? What would be the validity?