CS614 Quiz No. 2 Solution and Discussion

zareen

In context of data parallelism, to get a speed-up of N with N partitions, it must be ensured that:

There are enough computing resources
Query-coordinator is very fast as compared to query servers
Work done in each partition almost same
All of the given options

zareen

Mining multi dimensional databases allow users to:

Categorize the data
Analyze the data (Answer)
Summarize the data
All of the given options

zareen

A ______ index, if fits in the memory, costs only one disk I/O access to locate a record given a key.

zareen

Implementation of a data warehouse requires ________ activities

Highly integrated
Loosely integrated
Tightly decoupled
None of the given

zareen

In ________learning you don’t know the number of clusters and no idea about their attributes.

Supervised learning
Unsupervised learning (Answer)
Multi Dimension modeling
None of the given options

zareen

The exact formula for Speed-up is:

zareen

Spiral model is ________

zareen

An effective user education program includes, among others, the following guideline(s):

zaasmi

Question # 1 of 10 ( Start time: 10:37:17 PM ) Total Marks: 1
In context of data mining definition, the term “nontrivial” means:
Select correct option:
Discovering information is a simple task
Discovering information is a complex task (Correct)
We can not discover information
We simply find things rather than discovery

Question # 2 of 10 ( Start time: 10:37:38 PM ) Total Marks: 1
Identify the TRUE statement:
Select correct option:
Clustering is unsupervised learning and classification is supervised learning (Correct)
Clustering is supervised learning and classification is unsupervised learning
Both clustering and classification are unsupervised learning
Both clustering and classification are supervised learning

Question # 3 of 10 ( Start time: 10:38:00 PM ) Total Marks: 1
Waterfall model is appropriate when
Select correct option:
When the budget is low
When the deadline is strict
When resources are limited
Requirements are clearly defined (Correct)

Question # 4 of 10 ( Start time: 10:38:22 PM ) Total Marks: 1
Which of the following is the most ignored step during data warehouse development
Select correct option:
The requirement verification
The vision definition
Schema validation
Success criteria development (Correct)

Question # 5 of 10 ( Start time: 10:38:58 PM ) Total Marks: 1
In ______ phase of a fundamental data warehouse life cycle model, a working model of data warehouse is deployed for a selective set of users
Select correct option:
Design
Prototype (Correct)
Deployment
Operation

Question # 6 of 10 ( Start time: 10:39:19 PM ) Total Marks: 1
One of the drawbacks of waterfall model is that:
Select correct option:
Customers can not review the product during development
It does not work when the resources are limited
It does not define the project timeline/schedule
All of the given options (Correct)

Question # 7 of 10 ( Start time: 10:39:38 PM ) Total Marks: 1
Identify the TRUE statement:
Select correct option:
The data value increases as volume decreases (Correct)
The data value decreases as the volume decreases
The data value is independent of data volume
All of the given options

Question # 8 of 10 ( Start time: 10:39:56 PM ) Total Marks: 1
In contrast to statistics, data mining is ______ driven.
Select correct option:
Assumption
Knowledge (Correct)
Human
Database

Question # 9 of 10 ( Start time: 10:40:48 PM ) Total Marks: 1
In context of data mining definition, the term “value” means:
Select correct option:
The primary key of table
The index location of the record
Importance of hidden patterns discovered (Correct)
Numerical or string measure assigned to an attribute

Question # 10 of 10 ( Start time: 10:41:24 PM ) Total Marks: 1
In context of clustering, the term “distance” means:
Select correct option:
Similarity/dissimilarly of records (Correct)
The difference between the primary keys of two records
The relation of a record with corresponding record in child table
None of the given options

zareen

1_ In context of data parallelism, the work done by query processor should be:

Maximum

2_ _______ do not (typically) keep the index values in stored order

Hash based index

3_ if every key in the data is represented in the index file then it is called

Dense index

4_ In context of bitmap index, the length of the bit vector is:

the number of records in the base table

5_ In context of joining tables, the join condition is specified in _____ clause.

WHERE

6_ A join is identified by multiple tables in the _____ clause.

From

7_ Parallelism can be exploited, if there is:

All of the given options

8_ In ____ index, the ith bit is set to “1” if the ith row of the base table has the value for the indexed column.

Bitmap index

9_ As the number of processors increase, the speedup should also increase. thus we should have linear speedup. Which of the following is NOT the one of the barriers

to achieve this linear speed-up?

Amdah’l Law not sure

10_ Bitmap index is appropriate for:

Low cardinality data

Q1: in context of nested-loop join, actual number os matching rows returned as a result of the join would be ________ of the order of tables.
Independent.

Q2: Which of the following is NOT one of the parallel hardware architecture?
Shared Memory

Q3: If resources increase in proportion to increase in data size. time is constant’. The statement refers to:
Scale-Up

Q4: If every key in the data ﬁle is represented in the index ﬁle then it is called?
Dense Index

Q5: In context of data parallelism, to get a speed-up of N with N partitions, it must be ensured that.
All

Q6: In nested-loop join case, if there are ‘M’ rows in outer table and N rows in inner table, time complexity is.
o(MN)

Q7: The goal of__________ is to look at as few blocks as possib le to find the matching records(s).
Indexing

Q8: Parallelism can be exploited, if there is.
All of the given options

Q9: If we apply Run Length Encoding on the input “11001100”, the output will be.
21#20#21#20

Q10: Which of the following is NOT one of the variants of Nested-loop join?
Binary index nested-loop join.

Q11: In context of data parallelism, the work done by query processor should be:
Maximum.

Q12: ___________ do not (typically) keep the index values in sorted oreder
Hash based Index

Q13: if every key the data file is represented in the index file then it is called.
Dense Index

Q14: In context of bitmap index, the length of the bit vector is:
The number of records in the base table

Q15; In context of joining tables, the join condition is specified in ______ clause:
Where

Q16: A join is identified by multiple tables in the________ clause.
From

Q17: Parallelism can exploited, if there is
All of the given options

Q18: In ________ index, the ith bit is set to “1” if the ith row of the base table has the value for the index column
Bitmap index

Q19: As the number of processors increase, the speedup should also increase. Thus we should have linear speedup. Which of the following is NOT one of the barriers to achieve this linear speed-up?
Amdahl’ Law

Q20: Bitmap index is appropriate for:
Low cardinality data

Q21: If a task takes “T” time units to execute on a single data item, then execution of the task on “N” data items will take______ time units?
N*T

Q22: _________ lists each term in the collection only once and then shows a list of all the documents the contain the given term.
Inverted index

Q23: “More resources means proportionally less time for given amount of data”. The statement refers to:
Speed-UP

Q24: In context of data parallelism, to get a speed-up of N with N partitions, it must be ensured that:
All of the given option

Q25: In context of bitmap index, the length of the bit vector is
the number of records in the base table.

Q26: One of the preconditions to decide about operations to be parallelized is that:
Operation can be implemented independent of each other

Q27: A_________ index, if fits in the memory, costs only one disk I/O access to locate a record given a key.
Dense Index

Q28: In context of nested-loop join, actual number of matching rows returned as a result of the join would be ___ of the order of tables
Independent

Q29: __________ refers to “ Parallelexectution of single data operation across multiple partitions of data”
Data Parallelism.

A join is identified by multiple tables in the _ FROM ___ clause

In context of joining tables, the join condition is specified in _ WHERE ___ clause

The goal of ______ ing Goal _____ is to look at as few blocks as possible to find the matching records(s).

__ Sparse Index _____ index uses even less space than __ dense ____ index, but the block has to be searched, even for unsuccessful searches.

In context of data parallelism, to get a speed-up of N with N partitions, it must be ensured that:

If we apply Run Length Encoding on the input “11001100”, the output will be:

In B-tree index, the lowest level index blocks are called leaf blocks, and these blocks contain:

every indexed data value and a corresponding ROWID

___ Sparse Index ___ index stores first value in each block in the sequential file and a pointer to the block