Unit 6: Machine Learning AlgorithmsLinear Regression, KNN Classification & K-Means Clustering
The comprehensive CBSE Class 11 guide for Unit 6. Master Supervised Learning: Correlation, Pearson's r, Line of Best Fit (Least Squares), K-Nearest Neighbors (KNN), and Unsupervised Learning: 4 Clustering methods, K-Means mechanics, and 20 Handbook MCQs with Answer Key.
1. Correlation, Pearson's r & Linear Regression
Correlation measures the strength and direction of linear dependence between two continuous quantitative variables. It ranges from -1 to +1. Remember: "Causation is not always correlation".
Pearson's Correlation Coefficient Formula:
In the Handbook's age (x) vs. weight (y) calculation for 6 people (n = 6, Σx = 202, Σy = 409, Σxy = 13937, Σx² = 7280, Σy² = 28365):
r = 1004 / √8366284 ≈ 0.35 (positive moderate association).
Line of Best Fit & Least Squares Method:
Linear regression fits the line y = a + bx + e by minimizing the sum of squared differences between observed values and predicted values. The regression line always passes through the centroids of both features (μx, μy). In Excel, parameters are verified using =SLOPE() and =INTERCEPT().
2. Classification Paradigms & K-Nearest Neighbors (KNN)
2 class labels: Spam / Ham, Cancer / Normal, Pass / Fail.
> 2 mutually exclusive classes: Iris flower species, digits 0–9.
Instances belong to multiple classes (e.g. photo tags: apple + person).
Highly skewed majority vs. minority distribution (e.g. fraud detection).
Steps in the K-Nearest Neighbors (KNN) Algorithm:
- Select KKK: Choose the number of nearest neighbors (hyperparameter).
- Calculate Distance: Compute Euclidean distance from the query sample to all training points.
- Identify Neighbors: Select the KKK training instances with the smallest distances.
- Majority Vote: Count class frequencies among the KKK neighbors.
- Assign Class: Assign the candidate point to the class receiving the maximum vote.
3. Clustering Methods & K-Means Algorithm
1. Partitioning Clustering (K-Means)
Divides data into KKK non-hierarchical groups around centroids minimizing within-cluster variance.
2. Density-Based (DBSCAN)
Connects contiguous dense areas into arbitrarily shaped clusters separated by sparse regions.
3. Distribution-Based (GMM / EM)
Clusters data points based on probability distributions (e.g. Gaussian Mixture Models).
4. Hierarchical (Agglomerative)
Constructs a tree-like dendrogram that can be cut at any desired level without pre-specifying KKK.
The K-Means Clustering Algorithm Mechanics:
- • 1. Select KKK (number of clusters).
- • 2. Randomly place KKK centroids in the data space.
- • 3. Assign every data point to its closest centroid.
- • 4. Calculate the mean position of points in each cluster and update the centroid.
- • 5. Repeat reassignment and centroid recalculation until no points shift clusters (convergence).
Official Handbook MCQs & Answer Key
Click any option to instantly see if you're correct with the official CBSE explanation.
Which of the following are recognized types of correlation in statistical analysis?
Which technique analyzes the relationship between dependent and independent variables to provide a predictive mechanism?
Which visual plot is ideal for testing the linear relationship between an independent and a dependent variable?
Which of the following scatter plot patterns represents a positive correlation?
Which regression technique is used when there is only ONE independent variable predicting the dependent variable?
What is one major advantage of Linear Regression analysis?
What is Supervised Learning in Artificial Intelligence?
Which type of classification involves categorizing data into exactly TWO distinct classes?
What is Logistic Regression commonly used for in machine learning?
What is the primary goal of classification algorithms in AI?
The K-Nearest Neighbors (KNN) algorithm assigns a class label to a new data point by considering:
What is Clustering in Machine Learning?
Which category of machine learning does Clustering belong to?
Which clustering technique connects contiguous high-density regions separated by sparser areas?
The K-Means clustering algorithm is a classic example of which clustering category?
Which clustering approach generates a tree-like dendrogram that can be cut at different levels?
What is a recognized limitation of the K-Means clustering algorithm?
Which clustering method allows a single data instance to belong to multiple clusters with degrees of membership?
In the K-Means algorithm, what must be pre-specified by the user before running the clustering process?
What are data points that deviate significantly from the overall dataset pattern called?
Unit 6 Quick Recall Cheat Sheet
Frequently Asked Questions (FAQ)
Master Machine Learning Algorithms with 1:1 Live Tutoring
Understand the mathematical proofs and Python implementation behind Linear Regression, KNN, and K-Means. Get personalized 1-on-1 coaching to build practical ML workflows and score top marks in Class 11 AI.