Module 4-P1_and_P2 Quiz

K-means clustering is an example of:

Unsupervised learning (because it operates on data that has no predefined labels or targets)

Clustering groups data based on:

Similarity

K-means works best with:

Numerical data

Distance commonly used in K-means:

Euclidean

Centroid represents:

Cluster center (average/mean of all points within a cluster)

K-means output includes:

Cluster assignments

K-means is sensitive to:

Initialization (initial placement of centroids

K must be:

Predefined

WSS (Within-Cluster Sum of Squares) measures:

Cluster compactness

Elbow method helps:

Choose K

Association rules are used to:

Find relationships or patterns within large datasets

Apriori algorithm (association rule learning) works on:

Itemsets

Support measures:

Frequency (how frequently an itemset appears in a data set)

Confidence measures:

Conditional probability (if this then that)

Lift measures:

Independence

Frequent itemset means:

Appears often

Apriori property states:

Subsets are frequent

(all non-empty subsets of a frequent itemset must also be frequent.)

Classification is:

Supervised learning

Regression predicts:

Continuous Values

Logistic regression predicts:

Probability

Naive Bayes is:

a Classifier

(a supervised machine learning algorithm commonly used for classification tasks like spam filtering, document categorization, and sentiment analysis)

Decision tree is:

a Classifier

(to predict categorical target variables.)

Time series analysis studies:

Data over time

Text analysis uses:

TF-IDF (Term Frequency-Inverse Document Frequency)

Data analytics lifecycle includes:

Model planning

Model planning determines:

Method choice (the type of model/algorithm to be used)

Clustering does NOT:

Predict labels

Association rules example:

Market basket

Support threshold filters:

Frequent itemsets

(Filters out itemsets that are not frequent, keeping only those that meet the minimum support level.)

Confidence threshold filters:

Rules

(filter association rules, keeping only those that meet a minimum confidence level.)

Lift > 1 indicates:

Positive association

Leverage measures:

Difference

(difference between the observed co-occurrence of items and what would be expected if they were independent.)

K-means assumes clusters are:

Round

K-means limitation:

Not for categorical data

Cluster separation evaluated by:

Visualization

Outliers affect:

K-means strongly (because it uses the mean to calculate cluster centroids)

Apriori grows itemsets:

Iteratively (finds 1, uses 1 for 2, etc)

Time series detects:

Trends and seasonality

Text representation includes:

Bag of words (BoW): converts text into numerical vectors

TF-IDF (Term Frequency-Inverse Document Frequency) measures:

Term importance (evaluate how important or relevant a word is to a document in a collection or corpus)