BSCS 4-3 - Elective 2 (Machine Learning) - 5-4-2024

You are using the K-Nearest Neighbor algorithm to classify images of animals into different categories. If a new image of a cat is presented to the algorithm, and the 5 nearest neighbors are images of dogs, what class will the KNN algorithm assign to the new image?

• Cat

• Dog

• Misclassification

• None of the mentioned

Given a dataset where similar data points are clustered together, and you are required to categorize a newly introduced sample, which algorithm would be most appropriate to employ for this prediction task?

• Linear Regression

• K-Nearest Neighbors

• Support Vector Machines

• Naive Bayes Algorithm

Before proceeding with the K-Nearest Neighbor algorithm, what essential step should you take if you lack familiarity with the methods used to calculate the distance between points on a graph?

• Read about the Manhattan distance

• Thoroughly read about "Distance Between 2 Points."

• Study the mode calculation

• Learn about data sorting techniques

A data scientist is preparing to use K-Nearest Neighbor for a classification problem and has just imported a new dataset. What is the next step in the KNN procedure that should be undertaken?

• Determine the value of K

• Calculate distances between all points

• Load the data into the processing tool

• Normalize the dataset

A market researcher is using K-Nearest Neighbor to forecast customer preferences, and they have already pre-processed the dataset. What should their next immediate step in the KNN process?

• Choose the number of neighbors, K, to consider

• Randomly select data points as centroids

• Create a scatter plot of the data

• Implement cross-validation

Which of the following is not a step in the K-Nearest Neighbor algorithm?

• Load the data

• Initialize K to the chosen number of neighbors

• Calculate the mean of the K nearest neighbors

• Sort the distances from the query point to the training data points

What is the purpose of the "K" parameter in the K-Nearest Neighbor algorithm?

• It determines the number of nearest neighbors to consider

• It determines the distance metric to use

• It determines the number of features to use

• It determines the number of classes in the dataset

What is the MAIN advantage of the K-Nearest Neighbor algorithm?

• It can handle high-dimensional data

• It is a parametric algorithm

• It is computationally efficient

• It is easy to interpret the model

What implications does it have for the K-Nearest Neighbors algorithm to be described as a non-parametric model, particularly in terms of its assumptions about the structure and distribution of the data it processes?

• It does not assume anything about the underlying data distribution

• It requires a fixed number of parameters determined before training

• It forms clusters based on predefined parameters

• It can only be used with parametric datasets

Which types of problems is K-Nearest Neighbor primarily used to address?

• Clustering

• Classification

• Clustering and Classification

• Classification and Regression