Skip to content

R4HUL-ROY/GUI_DecisionTreeClassifier

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 

Repository files navigation

GUI_DecisionTreeClassifier

Provided features :

Upload Dataset : Here User can upload dataset by clicking the upload button. The checkbox above the upload button is to be checked if your dataset already have column/attribute name. In case your dataset does not contain any attribute name, you can unchecked the checkbox and this will create dummy column name inside automatically to avoid any unwanted errors.

Encoding : Here it use label encoding in case the value of an attribute is categorical user can select the multiple features at once to label encode them

Train model : By clicking train model button you can train and build your own dicision tree. Make sure you alreay upload a valid dataset before clicking on train button.

Predict : Predict button just simply evaluate the decision tree model with a test dataset and show the result on the console

Plot_tree : Plot tree button plot the decision tree on the console after training (Make sure to upload and train your model first before plotting)

Compare Result : Compare result button compare the results using Gini and Entropy as a splitting criteria. and show the result in each case side by side for better judgement

Parameters :

Train_test_split : Through Train test split we can set how much % we can take as a test set As example : Train_test_split = 0.2 means we want 20% data as our test data

Random state : Random state to fixed the psudo random number genereated each time when we run the program so that our result persists same each time we run, unless we change the random state parameter

Partitioning method : You can choose between two partiotioning method (Gini or Entropy) to build your decision tree model random state and partitioning method (gini/entropy) also through this GUI.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published