6. Supervised Learning – Linear Regression

🧠 Theory: In Supervised Learning, the model is trained on a labeled dataset, where both input features and the correct output are provided. Example: You give a model: ● Input: Years of Experience ● Output: Salary The model learns to predict salary from experience.

In supervised learning, we teach a model to make predictions based on examples we provide. The dataset used for training contains both the inputs (features) and the corresponding outputs (labels). For instance, if you want to predict a person's salary based on their years of experience, the years of experience act as the input while the salary is the expected output. Here, the model learns a relationship so it can predict the salary for new instances it hasn't seen before.

Imagine you are learning to bake. Your recipe (input) tells you the ingredients and the steps (features), while the baked cake (output) is the result. Just like in supervised learning, you follow the recipe to achieve the desired cake, the model learns from provided data to make predictions.

🧠 Theory: Linear Regression is a supervised learning algorithm that models the relationship between one dependent variable (target) and one or more independent variables (features) using a straight line. For Simple Linear Regression: y=mx+c ● y: dependent variable (e.g., salary) ● x: independent variable (e.g., experience) ● m: slope (coefficient) ● c: intercept (bias) 📈 Objective: Find the best-fitting straight line through the data that minimizes the prediction error.

Linear Regression seeks to find the best-fitting line that represents the relationship between the dependent variable (what we want to predict, like salary) and independent variable(s) (the input, like years of experience). The equation 'y = mx + c' describes this line where 'm' represents how steep the line is (slope), indicating how much 'y' changes for a change in 'x', and 'c' is where the line crosses the y-axis (intercept). The goal is to minimize the errors in our predictions by adjusting 'm' and 'c.'

Picture a line drawn across a scatterplot of points showing the relationship between study hours and test scores. Linear regression helps us find the best straight line through these points, predicting test scores based on study hours, much like how a coach designs a training plan for athletes to optimize performance.

Let’s create a small dataset: Years of Experience vs Salary import pandas as pd data = { 'Experience': [1, 2, 3, 4, 5], 'Salary': [35000, 40000, 50000, 55000, 60000] } df = pd.DataFrame(data) print(df)

We first need to create a dataset to work with. In this example, we construct a simple dataset using Python's pandas library. The dataset consists of two columns: 'Experience' representing years of work experience and 'Salary' representing the corresponding salaries. By putting this data into a DataFrame, we prepare it for analysis and modeling.

Think of this dataset like a box of LEGO bricks, where each brick represents a piece of information. Just as you can build various structures with LEGOs, we can use this dataset to draw insights, analyze trends, and build a prediction model about salaries based on experience.

Before training the model, let’s plot it: import matplotlib.pyplot as plt plt.scatter(df['Experience'], df['Salary'], color='blue') plt.xlabel('Years of Experience') plt.ylabel('Salary') plt.title('Experience vs Salary') plt.grid(True) plt.show()

Visualizing data helps us understand the relationship between variables before modeling. Here, we use matplotlib to create a scatter plot that displays each point representing a person's years of experience versus their salary. This visual representation can reveal patterns and trends in the data, guiding us in how we model it.

Imagine you are looking at a map that shows the locations of all the coffee shops around you. By visualizing this on a map, you can easily see where the clusters of coffee shops are located. In the same way, our scatter plot gives us a 'map' of how experience relates to salaries, helping us identify any trends.

We’ll use scikit-learn to build our model. from sklearn.linear_model import LinearRegression X = df[['Experience']] # Features y = df['Salary'] # Target model = LinearRegression() model.fit(X, y)

To build a machine learning model using linear regression, we leverage the scikit-learn library, which provides tools to easily implement models. In this chunk, we define our features (X) as the 'Experience' column and the target (y) as the 'Salary' column. We then create a Linear Regression model and train it with the existing data using the 'fit' method, allowing it to learn the relationship between experience and salary.

Consider teaching a child how to ride a bike. You hold onto the bike as they learn, offering guidance and support. Similarly, fitting our model with data allows it to learn the right relationship between inputs and outputs so it'll be able to predict future outcomes independently.

print("Slope (m):", model.coef_[0]) print("Intercept (c):", model.intercept_) If output is: Slope: 6250.0 Intercept: 28750.0 Then the model equation becomes: Salary=6250×Experience+28750

After training the model, we can extract the slope and intercept values, which are essential components of the regression equation. The slope indicates how much the salary is expected to increase with each additional year of experience, while the intercept represents the base salary when experience is zero. In our example, a slope of 6250 suggests that for every year of experience, the salary increases by $6250, while an intercept of 28750 indicates that the starting salary (with no experience) would be$28750.

Envision a climbing wall where each additional grip (year of experience) lets a climber go up higher (salary). The slope tells us how many meters higher the climber will go with each grip added. The intercept represents the view from the ground level, which in this analogy symbolizes the salary without any experience.

Let’s predict salary for 6 years of experience: predicted_salary = model.predict([[6]]) print("Predicted Salary for 6 years:", predicted_salary[0])

Once our model is trained, we can use it to make predictions. By inputting new data (in this case, 6 years of experience), we call the predict method to obtain an estimated salary based on the relationship the model learned from the training data. This enables us to see what salary the model predicts for a specific case.

Imagine you consult a financial advisor who uses historical salary data to estimate your future earnings based on your work experience. You provide your experience as input, and they give you an expected salary as an output, applying their expertise just like our model does.

plt.scatter(X, y, color='blue') plt.plot(X, model.predict(X), color='red') # Regression line plt.xlabel('Years of Experience') plt.ylabel('Salary') plt.title('Linear Regression') plt.show()

After making predictions, visualizing the regression line along with the data points provides clarity on how well the model fits the data. In this plot, the blue points represent actual salary data, while the red line shows the predicted relationship. The proximity of the data points to the line illustrates how accurately the model predicts salaries based on experience.

Think of it like drawing a line on a map that shows the route from a city where you start driving toward a destination. The actual roads (blue dots) might vary, but the ideal path (red line) shows the quickest way based on your knowledge. The regression line serves as the best path for predicting outcomes given new data.

Use Mean Squared Error (MSE) and R² Score: from sklearn.metrics import mean_squared_error, r2_score y_pred = model.predict(X) mse = mean_squared_error(y, y_pred) r2 = r2_score(y, y_pred) print("Mean Squared Error:", mse) print("R² Score:", r2) ● MSE: Lower is better ● R² Score: Closer to 1 is better (1 means perfect fit)

Evaluating the model's performance is crucial to understanding its accuracy. Mean Squared Error (MSE) quantifies how far off the predicted values are from the actual values; the lower the score, the better the model. The R² score indicates the proportion of variance in the dependent variable explained by the independent variables; a score closer to 1 indicates a better fit. Together, these metrics allow us to assess how well our model predicts salaries based on the years of experience provided.

Consider a student taking a test. The MSE is like calculating how many questions they answered incorrectly (the closer to zero, the fewer mistakes), while the R² score resembles their final grade, indicating how much they grasped the material. Both scores help us judge their performance in understanding the subject.

--