# This is how you place 3rd without probing ...

I started this competition because of it’s unique task. I liked the idea of „not overfitting“, very few data and it felt like some sort of „real world“ data science problem. Unfortunately I didn't have much time so I‘ve decided to wait and learn from others.
As it turns out lb probing was key for winning which was a little disappointing (though I really liked
[@zachmayer](/zachmayer)
&
[@cdeotte](/cdeotte)
solutions and findings!)
Now I have more time again so I’ve decided to climb lb and build a model without probing. 
Props to
[@featureblind](/featureblind)
&
[@nightwolfbrooks](/nightwolfbrooks)
- I used your models for some quick and dirty ensembling!
You can find my solution here
[link to kernel](https://www.kaggle.com/springmanndaniel/no-probing-autoenc-mlp-ensembled?scriptVersionId=14150758)
(feels like using a sledgehammer to crack a nut)
.
The key part is feature selection. I used „shadow“/ NULL – feature importance to select the final feature set
[link to kernel](https://www.kaggle.com/springmanndaniel/shadow-null-importance-feature-selection)
.
![feature selection 1](https://storage.googleapis.com/dbi-share/kaggle/feature_selection.png)
+Yes I know ... my code is really messy ;)
+It is always easier to climb lb after competition ended :)
+have fun!