Skip to main content
Cross Validated
Vedanshu's user avatar
Vedanshu's user avatar
Vedanshu's user avatar
Vedanshu
  • Member for 8 years, 2 months
  • Last seen more than a week ago
Profile Activity

35 Actions

All Accepts Posts Badges Comments Revisions Reviews Suggestions
awarded
 Notable Question
awarded
 Yearling
awarded
 Commentator
comment
Data balancing in image classification
@0asa I've to both segment and then classify.
asked
Data balancing in image classification
awarded
 Curious
awarded
 Informed
comment
Timeseries forecasting repeadted results
@SkanderH. It's univariate. Are hyperparameters responsibles for this ?
asked
Timeseries forecasting repeadted results
revised
Create score for comaparison of algorithms
edited tags
asked
Create score for comaparison of algorithms
awarded
 Popular Question
awarded
 Editor
revised
Testing a new model
area of topic reduced
asked
Testing a new model
comment
Proof of NOT initialization of weights to zero
@JanKukacka single hidden layer or may be a dense single layer at the output of a model.
comment
Proof of NOT initialization of weights to zero
Papers like arxiv.org/abs/1812.03425 shows that weight initialization to zero could lead to better results.
asked
Proof of NOT initialization of weights to zero
comment
MAPE changes when unnormalizing the data
I have seen many codes using lop1p for transforming the data, why would anyone want to do that then ?
comment
MAPE changes when unnormalizing the data
So, what's the solution ?
1
2