Virtual Labs

The parameter k in the KNN algorithm represents:

a: Number of clusters in the dataset Explanation

Explanation

b: Number of nearest neighbours used for classification Explanation

Explanation

c: Number of features in the dataset Explanation

Explanation

d: Number of training samples Explanation

Explanation

Which distance metric is most commonly used in KNN?

a: Manhattan distance Explanation

Explanation

b: Euclidean distance Explanation

Explanation

c: Cosine similarity Explanation

Explanation

d: Jaccard similarity Explanation

Explanation

How does the KNN algorithm determine the class of a new data point?

a: By calculating regression equations Explanation

Explanation

b: By majority voting among the nearest neighbours Explanation

Explanation

c: By clustering data points Explanation

Explanation

d: By calculating probability distributions Explanation

Explanation

If k = 1, classification is based on:

a: Majority voting Explanation

Explanation

b: Nearest neighbour Explanation

Explanation

c: Weighted average Explanation

Explanation

d: Regression Explanation

Explanation

The Iris dataset contains:

a: 100 samples Explanation

Explanation

b: 120 samples Explanation

Explanation

c: 150 samples Explanation

Explanation

d: 180 samples Explanation

Explanation

The KNN algorithm is considered a non-parametric model because:

a: It assumes the data follows a normal distribution Explanation

Explanation

b: It does not assume any predefined data distribution Explanation

Explanation

c: It estimates parameters using gradient descent Explanation

Explanation

d: It uses neural networks Explanation

Explanation

A very small value of k (for example k = 1) may lead to:

a: Underfitting Explanation

Explanation

b: High bias Explanation

Explanation

c: High sensitivity to noise Explanation

Explanation

d: Reduced computational cost Explanation

Explanation

When the number of features becomes very large, KNN suffers from:

a: Data imbalance Explanation

Explanation

b: Curse of dimensionality Explanation

Explanation

c: Gradient explosion Explanation

Explanation

d: Over-regularization Explanation

Explanation

The major computational disadvantage of KNN is:

a: Long training time Explanation

Explanation

b: High prediction time due to distance calculations Explanation

Explanation

c: Requirement of labelled data Explanation

Explanation

d: Difficulty handling small datasets Explanation

Explanation

KNN can be used for:

a: Only classification Explanation

Explanation

b: Only regression Explanation

Explanation

c: Both classification and regression Explanation

Explanation

d: Unsupervised clustering Explanation

Explanation

k-Nearest Neighbors (KNN)