How to Predict House Prices Using Machine Learning and Python

In this tutorial, we’ll use Python and a machine learning technique called Linear Regression to predict house prices based on some key features like the number of bedrooms, area, and crime rate in the area. Let’s break this down into easy steps.

1. Importing Libraries

First, we need some tools (libraries) to help us:

import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression
from sklearn.metrics import mean_squared_error
from sklearn.metrics import mean_absolute_error

pandas helps us manage our data.
train_test_split divides the data into training and testing sets.
LinearRegression creates the model that will predict house prices.
mean_squared_error and mean_absolute_error check how good our model is.

2. Loading the Data

We load our dataset (let’s assume it’s a file called Housing.csv) with information about houses:

data = pd.read_csv("Housing.csv")

Our data might look like this:

3. Defining Inputs and Output

Next, we pick what features we’ll use to predict the price (called X), and our target (price) is y.

X = data[['bedroom', 'area', 'crime']]  # The features we’ll use
y = data['price']  # The price we want to predict

4. Splitting the Data

We split the data into two parts:

• Training set: Used to train the model (80% of the data).
• Test set: Used to test how well the model performs (20%).

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

5. Building the Model

We create a simple Linear Regression model:

model = LinearRegression()

6. Training the Model

Now we train the model using the training data:

model.fit(X_train, y_train)

7. Making Predictions

After training, we use the model to predict house prices from our test data:

y_pred = model.predict(X_test)

8. Checking the Model’s Accuracy

We check how good our predictions are using two common methods:

Mean Squared Error (MSE): Squared average of errors (again, lower is better).

Mean Absolute Error (MAE): Average of the errors (lower is better).

mae = mean_absolute_error(y_test, y_pred)
mse = mean_squared_error(y_test, y_pred)
print(f"MAE: {mae}")
print(f"MSE: {mse}")

9. Predicting a New House Price

Finally, we can use the model to predict the price of a new house. Let’s say the house has 3 bedrooms, 700 sqft of area, and a crime rate of 1:

my_predict = model.predict(pd.DataFrame({'bedroom': [3], 'area': [700], 'crime' : [1]}))
print(f"Price: ${my_predict[0]:,.2f}")

If everything is set up correctly, it will predict the price based on those values.

Complete Code:

import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression
from sklearn.metrics import mean_squared_error
from sklearn.metrics import mean_absolute_error

data = pd.read_csv("Housing.csv")
#data.head()

X = data[['bedroom', 'area', 'crime']]
y = data['price']

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
model = LinearRegression()

model.fit(X_train, y_train)
y_pred = model.predict(X_test)

mae = mean_absolute_error(y_test, y_pred)
mse = mean_squared_error(y_test, y_pred)
print(f"MAE: {mae}")
print(f"MSE: {mse}")

my_predict = model.predict(pd.DataFrame({'bedroom': [3], 'area': [300], 'crime' : [1]}))
print(f"Price: ${my_predict[0]:,.2f}")