(оригинал: Texas - Summer son)
It could be system
With too much
Computing elements.
Where is the difference due to
Decaying gradients?
Before you change the weights, reconsider.
Before you add new layer, reconsider.
Training needs testing,
Cross-validating.
We write the finite sum
To solve least squares.
Deriving it again
Passed on to you.
The model can compute something
But doesn’t explain us
What mechanism is revealed
By data samples.
Before you use results, reconsider.
Before interpreting them reconsider.
It may show nothing
And may cause laughing.
We write the finite sum
To solve least squares.
Deriving it again
Passed on to you.
May have no meaning
That you are seeing.
We write the finite sum
To solve least squares.
Deriving it again
Passed on to you.
Finite sum
© Copyright: Глеб Гренкин, 2023
Свидетельство о публикации №123082802409
Свидетельство о публикации №123082802409
Рецензии