2.2. Supervised Learning — Learning with Answers

“Supervised” means the computer learns from example problems that already have the correct answers. Imagine you’re learning to predict student marks based on study hours. You get a chart:

Hours Studied 1 35 2 45 3 55 You learn from this pattern. That’s exactly what a computer does.

Supervised Learning is a process where a computer is trained using data that comes with the answers already known. For example, when trying to predict a student's marks based on how many hours they studied, the model uses existing data, like a table showing study hours and corresponding marks, to learn the relationship between these variables. The computer analyzes this data to discover patterns which it can use later to make predictions for new, unseen data.

Think of it like a student preparing for an exam using a study guide. The guide has questions and the correct answers listed, allowing the student to learn through examples. By studying these, they can better understand the subject and perform well on similar questions during the actual exam.

Task Input Output Predicting house prices Area, Location, Bedrooms Price Email spam detection Words in email Spam or Not Spam Disease diagnosis Patient details Has disease or not Credit risk Age, Salary, History Safe or Risky

Supervised Learning is widely applied in various fields due to its ability to make accurate predictions. For instance, in real estate, inputs like the area, location, and number of bedrooms can be fed into a model to predict house prices. Similarly, email spam detection uses specific words in the email to determine if it is spam or legitimate. Medical fields utilize this method to diagnose diseases based on patient details, while financial institutions assess credit risk through factors like age, salary, and credit history.

Consider a doctor diagnosing patients. The doctor uses information from many patients, such as symptoms and test results, to learn which conditions are associated with which symptoms. The next time they see a new patient, they can use what they've learned to make a diagnosis.

1. Regression — Output is a number E.g., Predict marks, temperature, price
2. Classification — Output is a category E.g., Spam or Not Spam, Pass or Fail

Supervised Learning can be divided into two main types: Regression and Classification. Regression is used when the output is a continuous value. For example, predicting a student's marks based on hours studied generates a numerical result. In contrast, Classification is used when the output falls into categories; for example, determining whether an email is 'spam' or 'not spam' is a classification problem.

Imagine trying to determine the weather for the coming week. If you guess the temperature using past temperatures, that's regression—you're predicting a specific number. However, if you classify the weather as 'sunny' or 'rainy,' that falls under classification—you're assigning it to a category.

Let’s predict marks from hours studied. from sklearn.linear_model import LinearRegression import numpy as np X = np.array([[1], [2], [3], [4], [5]]) # Hours y = np.array([35, 45, 55, 65, 75]) # Marks model = LinearRegression() model.fit(X, y) print("Prediction for 6 hours:", model.predict([[6]])[0])

This example demonstrates how to implement a regression model to predict marks based on study hours. The model analyzes the relationship between the number of hours studied (input) and the corresponding marks (output). It fits a line to the data points to find the best representation of the trend. When we input 6 hours into the model, it predicts what the marks will be using the learned relationship.

Think of this like plotting how your performance improves as you practice a sport. If you track how many hours you practice and your corresponding scores or wins, you could draw a line that best explains how practice correlates with success.

Now let’s predict pass/fail: from sklearn.neighbors import KNeighborsClassifier import numpy as np X = np.array([[1], [2], [3], [4], [5]]) y = np.array([0, 0, 1, 1, 1]) # 0 = Fail, 1 = Pass model = KNeighborsClassifier(n_neighbors=3) model.fit(X, y) print("Prediction for 2.5 hours:", model.predict([[2.5]])[0])

In this example, we use a classification model (K-Nearest Neighbors) to predict whether a student passes or fails based on study hours. The model looks at the study hours of other students and their results to determine the outcome for a new student who studied for 2.5 hours. It checks the results of the closest three peers to make a prediction.

Imagine assessing whether a friend will enjoy a movie. If you know they liked similar movies, you can predict they would also like the new one based on those closest matches. This is similar to how classification models work—they learn from examples to categorize new cases.

--