Metadata-Version: 2.1
Name: bianary-decision-parameter
Version: 0.2
Summary: bianary_classification
Home-page: UNKNOWN
Author: MAINAK RAY
Author-email: mainakr748@gmail.com
License: MIT
Keywords: decision
Platform: UNKNOWN
Classifier: Development Status :: 5 - Production/Stable
Classifier: Intended Audience :: Education
Classifier: Operating System :: Microsoft :: Windows :: Windows 10
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3.8
Description-Content-Type: text/markdown
Requires-Dist: sklearn
Requires-Dist: pandas
Requires-Dist: numpy

this module is used to find out best parameters of desision tree.
![Markdown Logo](https://github.com/mainak-cmd/All-Ml/blob/main/flow_chart.drawio.png)

#Bianary_classification

##Aim
The module aims to create a library that will help users to build trees that can help classify binary classes easily. But, it is essential to note that this module should be used after the process (cleaning, transformation, and reduction) of the data.

##Advantages
1.)Autopruneing the tree no need to hyperparameter tuneing
2.)very easy to use.
3.) Brier score to evaluate to find out best paraameter as  is that it is focused on the positive class, which for imbalanced classification is the minority class. A perfect classifier has a Brier score of 0.0.

##Limitation
1.) Preprocessing must requires.
2.) do not shows good perfformance for high dimentional data ,feature reduction before use this module is adviceable.
3.)little time consumeing.


##code example : 
```python
import bianary_decision_parameter
from bianary_decision_parameter import 
from sklearn import preprocessing
from sklearn.datasets import load_breast_cancer
from sklearn.model_selection import train_test_split
Load breast_cancer dataset
split it into train and test set
da=load_breast_cancer()
X1=da.data
y1=da.target
scaled_X_1 = preprocessing.scale(X1)
X_train1, X_test1, y_train1, y_test1 = train_test_split(scaled_X_1, y1,test_size=0.3)
p2=bianary_decision_parameter('auto',X_train1,y_train1, X_test1, y_test1,1,1)
p3=bianary_decision_parameter('sqrt',X_train1,y_train1, X_test1, y_test1,1,10) 
p4=bianary_decision_parameter('log2',X_train1,y_train1, X_test1, y_test1,1,10)

p2.parameter()

	accuracy_score  missclassification 	specificity 	Recall 	precision 	depth 	ccp_alpha 	max_features
     0 	        0.912281 	0.087719 	0.912281 	0.83871 	0.912281 	2 	0.004307 	auto
```
## Explanation
p2=bianary_decision_parameter('auto',X_train1,y_train1, X_test1, y_test1,1,1)
Feature_selection = 'auto' \n
class_weight = 0_class ,1_class



