Polynomial Regression - helpcodes.me

Part 3:  Polynomial Regression : ( Python )

 

Today we will learn more about the Polynomial Regression and which regression is better Linear or Polynomial . The Polynomial vs Linear Regression ? , Polynomial key factor and uses. 

Importing data and Liberaries

import numpy as np

import matplotlib.pyplot as plt

import pandas as pd

# Importing the dataset

dataset = pd.read_csv('position_salaries.csv')

X = dataset.iloc[:, 1:2].values

y = dataset.iloc[:, 2].values

Now we will Splitting the data in train and test data sets 

# Splitting the dataset into the Training set and Test set

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=0)

Now we will do Linear regression using the Sklearn liberaries inwhich we used LinearReression Class and ' lreg ' Objects

# Fitting Linear Regression to the dataset

from sklearn.linear_model import LinearRegression

lreg = LinearRegression()

lreg.fit(X, y)

After Fitting data in Linear model now we will go Toward Displaying thegraph of Linear Regression and try to figure out in visulisation method 

def vlinear():

    plt.scatter(X, y, color='red')

    plt.plot(X, lreg.predict(X), color='blue')

    plt.title('Truth or Bluff (Linear Regression)')

    plt.xlabel('Position level')

    plt.ylabel('Salary')

    plt.show()

    return

vlinear()

After Completeing the Linear Model now we move forward to ward the Polynomial Model 

The equation of Polynomial Regression is 

            Y = β + β  . X1 + β . X1^2 + β . X2^3  + β . X3^4

 we need the Square term of each Coolum to get the Best prediction and good result .

Now We fill Fit The Data in PolyNomial Regression model :

# Fitting Polynomial Regression to the dataset

from sklearn.preprocessing import PolynomialFeatures

poly_reg = PolynomialFeatures(degree=4)

X_poly = poly_reg.fit_transform(X)

pol_reg = LinearRegression()

pol_reg.fit(X_poly, y)

Now We Fit  The data In Regression model and now we will Visulize the data Using the sklearn libararies and we will compare the linear and Polynomial regression and the output of both are give below in screen shots.

def vpolymonial():

    plt.scatter(X, y, color='red')

    plt.plot(X, pol_reg.predict(poly_reg.fit_transform(X)), color='blue')

    plt.title('Truth or Bluff (Linear Regression)')

    plt.xlabel('Position level')

    plt.ylabel('Salary')

    plt.show()

    return

vpolymonial()

Comparison :

The Polynomial is accurate than the linear and giving best prediction .The polynomial is very closed to exact result and data .