Jawaharlal Nehru Technological University Hyderabad
Jawaharlal Nehru Technological University Hyderabad MCA 2018 DATA WAREHOUSING AND DATA MINING Previous Question Paper (June/July - 2018 - R13)
Part A is compulsory which carries 20 marks. Answer all questions in Part A. Part B consists of 5 Units. Answer any one full question from each unit. Each question carries 8 marks.
Part A
What are the characteristics of an interesting pattern?
Answer
A pattern is considered interesting if it is easily understood by humans, valid on new data with some degree of certainty, potentially useful, and novel or validates a hypothesis that the user sought to confirm. Furthermore, interesting patterns should be actionable, meaning they can lead to specific decisions or actions that provide value to the user.
What is meant by multi dimensional data model?
Answer
A multidimensional data model is a conceptual framework used to organize and represent data in a way that facilitates complex analytical queries. It typically organizes data into dimensions (attributes or perspectives like time, location, or product) and measures (quantitative values like sales or revenue). This model is often visualized as a data cube, allowing users to perform operations such as roll-up, drill-down, slice, and dice to view data from different angles.
Give examples for a single dimensional association rule and a quantitative multidimensional association rules.
Answer
Sign in to view the complete answer paper.
What are the accuracy measures for a classifier?
Answer
Sign in to view the complete answer paper.
List the merits and demerits of hierarchical agglomerative clustering.
Answer
Sign in to view the complete answer paper.
Part B
What is data mining? Explain it as a step in knowledge discovery process.
Answer
Sign in to view the complete answer paper.
Demonstrate attribute subset selection as a preprocessing technique.
Answer
Sign in to view the complete answer paper.
Define data warehouse. Compare it with database management systems.
Answer
Sign in to view the complete answer paper.
Explain BUC algorithm for data cube computation.
Answer
Sign in to view the complete answer paper.
Using FP Growth algorithm find frequent item sets (support threshold 30%) for the provided data.
Answer
Sign in to view the complete answer paper.
What is correlation analysis? Explain the significance of lift measure for correlation analysis.
Answer
Sign in to view the complete answer paper.
How to prepare data for classification? Explain with suitable data set.
Answer
Sign in to view the complete answer paper.
What are the characteristics of neural network that make a good classifier? Describe back propagation algorithm.
Answer
Sign in to view the complete answer paper.
Explain k-means algorithm and contrast it with k-medoid algorithm.
Answer
Sign in to view the complete answer paper.
What is an outlier? What is the need of outlier detection? Explain any one technique for outlier analysis.
Answer
Sign in to view the complete answer paper.