Python Breast Cancer Prediction using Random Forest
PYTHON BREAST CANCER PREDICTION
USING RANDOM FOREST
Python Breast Cancer prediction is a simple project in python which is used to classify
whether the person is having breast cancer or not i.e. by classifying the given data into
malignant or benign. This project is built using Random Forest classifier i.e. the model is
trained using Random Forest and achieved an accuracy of 97%. If the output is malignant
then the person is suffering from breast cancer, if not then the person does not have breast
cancer.
In [88]:
import pandas as pd
In [89]:
import numpy as np
In [90]:
import matplotlib.pyplot as plt
In [91]:
import seaborn as sns
In [92]:
from sklearn.datasets import load_breast_cancer
In [93]:
cancer=load_breast_cancer()
cancer
Out[93]:
In [94]:
cancer.keys()
Out[94]:
In [95]:
print(cancer["DESCR"])
#Data set description
In [96]:
print(cancer["target"])
#target values
#0 indicates malignant
#1 indicates benign
In [97]:
cancer["target_names"]
#target names
Out[97]:
In [98]:
cancer["feature_names"]
#feature names
Out[98]:
In [99]:
cancer["data"].shape
#data shape i.e 569 rows and 30 columns
Out[99]:
In [100]:
df_cancer=pd.DataFrame(np.c_[cancer["data"],cancer["target"]],columns=np.append(cancer["feature_names"],["target"]))
df_cancer
#data set after forming it into dataframe
Out[100]:
In [101]:
sns.pairplot(df_cancer,hue="target",vars=['mean smoothness','mean perimeter','mean texture','mean area','mean radius'])
#pairplots between 4 features mean radius,mean texture,mean perimeter , mean area
Out[101]:
In [102]:
sns.countplot(df_cancer["target"])
#histogram representing number of 0's and 1's in target
Out[102]:
In [103]:
plt.figure(figsize=(30,10))
sns.heatmap(df_cancer.corr(),annot=True)
#heat map representing corelations between each feature
Out[103]:
In [104]:
x=df_cancer.drop(["target"],axis=1)
x
Out[104]:
In [105]:
y=df_cancer["target"]
y
Out[105]:
In [106]:
from sklearn.model_selection import train_test_split
X_train,X_test,y_train,y_test=train_test_split(x,y,test_size=0.2,random_state=5)
#spliting data into test and train data
In [107]:
from sklearn.ensemble import RandomForestClassifier
RF_MODEL= RandomForestClassifier(n_estimators=100)
RF_MODEL=RF_MODEL.fit(X_train,y_train)
#Training model using Random Forest Classifier
In [108]:
from sklearn.metrics import classification_report
In [109]:
y_predict=RF_MODEL.predict(X_test)
In [110]:
y_predict
#predicted output by RF model
Out[110]:
In [113]:
print(classification_report(y_test,y_predict))
Comments