UBC-MDS / Breast-Cancer-Prediction

Supervised Learning Model to predict breast cancer
MIT License
1 stars 11 forks source link

Explanation of suitability of decision tree model to our analysis #12

Closed talhaadnan100 closed 5 years ago

talhaadnan100 commented 5 years ago

Incorporate advisor's feedback to Readme:

Why do you think decision tree model is suitable for your analysis?

talhaadnan100 commented 5 years ago

@akanshaVashisth the explanation for choosing decision tree classification is as follows:

We choose decision tree classification for our analysis because it is parametric. In our attempt to build a model that ranks features based on their importance, decision tree classification takes all of the features and complete training data to pick the strongest predictors. Other supervised learning approaches that are non-parametric such as K-Nearest Neighbours would not be able to rank the features by importance, and thus, fail to answer our analysis question.

Kindly choose the issue after reviewing.

akanshaVashisth commented 5 years ago

Great explanation!