To enhance evaluation metrics you still can:
- try some feature selection methods
- use more data (if possible)
On Wed, Mar 19, 2014 at 1:09 PM, James Bergstra wrote:
> If there are no hyper-parameters, then you don't need to do any such
> optimization :)
>
>
> On Wed, Mar 19, 2014 at 8:04 AM, A
If there are no hyper-parameters, then you don't need to do any such
optimization :)
On Wed, Mar 19, 2014 at 8:04 AM, Ahmed Ibrahim wrote:
> Hi,
>
> 1) Is it possible to optimise Gaussian Naive Bayes using GridSearch?
>
> 2) Since there are no parameters in GussianNB, do I have to use another
>
Hi,
1) Is it possible to optimise Gaussian Naive Bayes using GridSearch?
2) Since there are no parameters in GussianNB, do I have to use another
approach for optimisation?
Thanks
Ahmed
--
Learn Graph Databases - Downlo