Speaker
Tommaso Dorigo
(Universita e INFN, Padova (IT))
Description
We demonstrate how a nearest-neighbor algorithm can be endowed with a large number of free parameters by assigning weights and biases to all training events. The simultaneous optimization of the large number of parameters by gradient descent allows to obtain similar performances to those of neural networks or boosted decision trees, although at a much higher CPU price.
Details
Dr Tommaso Dorigo, INFN, Sezione di Padova, Italy
www.pd.infn.it
| Is this abstract from experiment? | No |
|---|---|
| Name of experiment and experimental site | N/A |
| Is the speaker for that presentation defined? | Yes |
| Internet talk | Maybe |
Author
Tommaso Dorigo
(Universita e INFN, Padova (IT))
