Sign up to take part
Registered users can ask their own questions, contribute to discussions, and be part of the Community!
Hello community. I was going through Machine Learning Quick Start,
but I don't understand the results/conclusion. Would someone assist me in understanding? What did this ML magic visualize in the end?
Thanks and my apologies. I am not a data scientist, just an old IT nerd.
In the tab you linked ("Inspect a model's results"), you can see three things :
1. What are the features that impacted the most the models prediction. Said differently, if you tried to "location" and "country" by the same "amount of difference", moving country would make the result change the most.
Of course, there's no common way to changing location and country, so what this really mean is that we compare changing the numerical values of those things after the "feature handling" part of your model, which converts them to numerical values.
2. The performance of the model. I think this is more explicit, you can see how your model perform among different criterias. Since it tries to differentiate two outputs, how often does it get output 1 correctly ? output 2 ? Does it tend to fail because it choses output 2 instead of 1, or the opposite ?
3. A reminder of how you handled your features, aka how you converted the bunch of data you had to begin with into only numerical values. This is useful to understand the 1st screen, for example maybe you'll realize one of your features isn't rescaled and that's why it became so important !
I hope this helps, and don't worry, I can attest there's no age limit to get into machine learning !