21 Aug 2024

Posted on:

12 Aug 2024

0

# Decision Trees - Prediction - I have issues with the probabilites values:

Dear Professor Kitov

In this section`s excel file i get different values for probabilites as it would make sense to get "0" value for the Rainy/windy situation as even the table shows do not train outside, I appreciate if you check this?

Regards,

Mani

Instructor
Posted on:

21 Aug 2024

1
Hello Mani,
Thanks for reaching out!
Based on the question you're asking, it seems there is some confusion or discrepancy regarding the probability values in the decision tree output, specifically for the "Rainy/Windy" situations. Let's clarify this by breaking it down.
In decision trees, the model makes decisions based on the input features. Once the model has enough data to confidently make a decision, it assigns a probability. In this table, the probability is 1 for every decision, indicating that the model is 100% confident about each decision based on the provided features.
The probability reflects the model’s confidence in its decision. A probability of 1 means the model is fully confident in the outcome, while a probability of 0 would indicate complete certainty that the decision is negative (e.g., "Do Not Train Outside").
The table you provided indicates that for rainy and windy conditions (Windy = TRUE and Outlook = Rainy), the decision is consistently "No" for "Train Outside." In these cases, it would make logical sense for the probability to reflect the model's confidence in not training outside. So, the probability would be 1-  indicating that the model is 100% confident about each decision based on the provided features.
I hope this makes sense!
Best,
The 365 Team