LAB MANUAL
Unit III – Machine Learning
Disclaimer: The content is curated from online/offline resources and used for educational purpose only
Disclaimer: The content is curated from online/offline resources and used for educational purpose only
Unit III – Machine Learning
Lab 2. Predicting Solar Power Output Using Linear Regression
Objective
To implement solar power output prediction using Linear Regression, a general approach
to how you can tackle the solar power output prediction problem using linear regression,
along with a sample implementation using Python and common libraries like Pandas,
Scikit-learn, and Matplotlib. The model will be trained on historical data and will predict
the solar power output for given values of the input features. The performance of the
model will be evaluated using standard metrics such as Mean Squared Error (MSE) and
R-squared (R²).
Problem
The global shift towards sustainable energy has led to an increasing reliance on
renewable energy sources such as solar power. Accurate predictions of solar power
output are crucial for effective grid management, energy storage, and optimizing the use
of solar energy. we aim to predict the solar power output based on certain environmental
factors, which include temperature, humidity, solar_irradiance and wind_speed. The goal
is to develop a predictive model using Linear Regression, a simple yet powerful algorithm,
to estimate the solar power output from these features.
Solution
To predict solar power output using linear regression and save the model, we'll go through
the following steps:
1. Import required libraries
2. Prepare the dataset
1. Load dataset
2. Select Features and target variable
3. Train a linear regression model
1. Split the dataset into training and testing sets (80% train, 20% test)
Disclaimer: The content is curated from online/offline resources and used for educational purpose only
2. Initialize and train the linear regression model
4. Evaluate the model
1. Predicting the solar power output on the test set
2. Evaluate the model performance
5. Visualization
6. Save the model for later use
7. Load and use the Model
Procedures
1. Import required libraries
2. Prepare the dataset
Features:
• temperature: in the range of 10-35°C to represent different climates.
• humidity: In between 20% and 100% as humidity varies across regions.
• solar_irradiance: In between 100 W/m² and 1000 W/m² (typical range for solar
radiation).
• wind_speed: In between 0 m/s and 10 m/s, representing wind conditions.
Target Variable (solar_power_output): This is calculated as a function of the other
features, with random noise added to make the data more realistic. The relationship
assumes that higher irradiance and temperature increase power output, while high
humidity reduces efficiency. Wind speed has a small positive impact.
Disclaimer: The content is curated from online/offline resources and used for educational purpose only
2.1 Load Dataset
2.2 Select Features and target variable
3. Train a linear regression model
Now, we’ll split the dataset into training and testing sets, train the linear regression model,
and evaluate its performance.
3.1 Split the dataset into training and testing sets (80% train, 20% test)
3.2 Initialize and train the linear regression model
Disclaimer: The content is curated from online/offline resources and used for educational purpose only
4. Evaluate the model
We will now predict the solar power output using the test data and evaluate the
performance using metrics like Mean Squared Error (MSE) and R-squared (R²).
4.1 Predicting the solar power output on the test set
Predicted outputs of test dataset,
Disclaimer: The content is curated from online/offline resources and used for educational purpose only
4.2 Evaluate the model performance
5. Visualization
Disclaimer: The content is curated from online/offline resources and used for educational purpose only
• Each blue dot represents a comparison between an actual and predicted value.
• The red dashed line represents a perfect fit line, where predicted values would
exactly match actual values if the model were perfect.
• The closer the points are to this line, the better the model's accuracy.
Disclaimer: The content is curated from online/offline resources and used for educational purpose only
• Plots both actual and predicted solar power outputs for each test data point index.
• Helps visualize how predictions follow (or deviate from) the actual trend across the
dataset.
6. Save the model for later use
Disclaimer: The content is curated from online/offline resources and used for educational purpose only
7. Load and use the Model
Disclaimer: The content is curated from online/offline resources and used for educational purpose only
