Using quadratic linear regression #1840

Onelinerhub · Dec 16, 2022 · 8a1f66c · 8a1f66c
1 parent 9e40069
commit 8a1f66c
Showing 1 changed file with 52 additions and 0 deletions.
diff --git a/python-scikit-learn/using-quadratic-linear-regression.md b/python-scikit-learn/using-quadratic-linear-regression.md
@@ -0,0  1,52 @@
 # Using quadratic linear regression
 
 ### Quadratic regression can be achieved by using `PolynomialFeatures` to prepare dataset for polynomial form:
 
 ```python
 from sklearn import preprocessing, linear_model
 import numpy as np
 
 data = np.array([[1, 1], [2, 4], [3, 9], [4, 15], [5, 21], [6, 36]])
 X = np.array(data[:,0]).reshape(-1,1)
 y = data[:,1]
 
 poly = preprocessing.PolynomialFeatures(degree=2, include_bias=False)
 Xp = poly.fit_transform(X)
 
 model = linear_model.LinearRegression()
 model.fit(Xp, y)
 
 y_pred = model.predict(Xp)
 ```
 
 - `from sklearn import` - import module from [lib:scikit-learn](https://onelinerhub.com/python-scikit-learn/how-to-install-scikit-learn-using-pip)
 - `.PolynomialFeatures(` - create polynomial object to prepare our dataset for polynomial form
 - `poly.fit_transform(` - transform original dataset to polynomial form
 - `linear_model.LinearRegression` - initialize linear regression model
 - `.fit(` - train transformation model
 - `.predict(` - predict target variable based on given features dataset
 
 group: linear
 
 ## Example: 
 ```python
 from sklearn import preprocessing, linear_model
 import numpy as np
 
 data = np.array([[1, 1], [2, 4], [3, 9], [4, 15], [5, 21], [6, 36]])
 X = np.array(data[:,0]).reshape(-1,1)
 y = data[:,1]
 
 poly = preprocessing.PolynomialFeatures(degree=2, include_bias=False)
 Xp = poly.fit_transform(X)
 
 model = linear_model.LinearRegression()
 model.fit(Xp, y)
 
 print(model.predict(Xp))
 ```
 ```
 [ 1.57142857  3.62857143  7.97142857 14.6        23.51428571 34.71428571]
 
 ```