data mining

Question 1

t/f: data mining has one definition

Answer 1

false, has many

Question 2

data mining

Answer 2

• non-trivial extraction of implicit, previously unknown and potentially usefully information from data
• exploration & analysis, by automatic or semi-automatic means, of large quantities of data in order to discover meaningful patterns

Question 3

origins of data mining

Answer 3

ideas come from many disciplines including machine learning/AI, pattern recognition, statistics, and database systems

Question 4

compared to data mining, traditional techniques may be unsuitable due to

Answer 4

• enormity of data
• high dimensionality of the data
• heterogeneous, distributed nature of data

Question 5

types of data mining algorithms

Answer 5

supervised algorithms (classification) and unsupervised algorithms (clustering)

Question 6

supervised algorithms

Answer 6

• learning by example
• use training data which has correct answers (class label attribute)
• create a model by running the algorithm on the training data
• identify a class label for the incoming new data

Question 7

unsupervised algorithms

Answer 7

• do not use training data
• classes may not be known in advance

Question 8

approaches for supervised (classification)

Answer 8

• decision trees
• regression
• neural networks
• support vector machines
• K-Nearest Neighbor approach
• Bayesian Classification

Question 9

classification: description

Answer 9

• given a collection of records. each record contains a set of attributes, one of the attributes is the dependent variable/class.
• find a model to predict the class attribute as a function of the values of the other attributes

Question 10

classification goal

Answer 10

previously unseen records should be assigned to a class as accurately as possible

Question 11

classification: a ____ is used to determine the accuracy of the model.

Answer 11

test set
- usually, the given data set is divided into training and test sets, with training set used to build the model and test set used to validate it

Question 12

classification approach: k-Nearest Neighbor

Answer 12

look at characteristics/attributes
“if it walks like a duck and quacks like a duck, then it’s probably a duck”

Question 13

nearest neighbor requires 3 things

Answer 13

• the set of stored records
• distance metric to compute the distance between records
• the value of k, the number of nearest neighbors to retrieve

Question 14

to classify an unknown record

Answer 14

• compute distance to other training records
• identify k nearest neighbors
• use class labels of nearest neighbors to determine the class label of unknown record

Question 15

choosing the value of k

Answer 15

if k is too small, model is sensitive to noise.
if k is too large, neighborhood may include too many points from other classes

Question 16

unsupervised algorithm: clustering

Answer 16

given a set of data points, each having a set of attributes, and a similarity measure among them, find clusters such that
- data points in one cluster are more similar to one another
- data points in separate clusters are less similar to one another

Question 17

clustering similarity measures

Answer 17

• euclidean distance (if attributes are continuous)
• other problem-specific measures

Question 18

examples of classification

Answer 18

• direct marketing: targeting a set of consumers likely to buy new product (use data for similar product introduced before)
• fraud detection: label past transactions as fair or fraud to form class attribute and train the model

Question 19

examples of clustering

Answer 19

• market segmentation: collect different attributes of customers, find clusters of similar customers (measure cluster quality by observing buying patterns of customers in same cluster vs those from diff clusters)
• document clustering: identify frequently occurring terms in each document, form a similarity measure based on the frequencies of different terms, and use it to cluster.