ITM402
Data
Warehousing & Data Mining
Assignment - II
Assignment
Code: 2016ITM402A2 Last Date
of Submission: 30th April 2016
Maximum
Marks: 100
Attempt all the questions.
All the questions are compulsory and carry equal marks.
Section-A
1 What do you know about decision
trees? How do the decision trees work? Discuss the CART algorithm.
2 What do you mean by Neural
Networks? How it works? How neural network techniques are applied in data
mining?
3 What is web mining? How does data
mining work? How it is used in business.
4 Explain in brief how the
information from a data warehouse promotes CRM (Customer Relationship Management
Section-B
Case Study
Neural
Networks have probably seen their greatest acceptance and application in the
financial industry. Applications such as credit card fraud, default (personal
bankruptcy), and even customer attrition have all shown successful application
with neural networks. For fraud alone, the dollar amounts to be saved by viable
predictive models are staggering. In 1995 the combined losses from credit card
fraud and counterfeiting was $1.3 billion. Visa member banks alone lost more
than $148 million to counterfeiters in 1994. The good news is that neural
network system have been introduced that reduced that loss by 16 % to $124
million just one year later.
Despite
successes like these in fraud and other applications, the holy grail of the
financial applications is still time series prediction: being able to say what
is going to happen next- whether it is predicting the closing price of a stock,
the market, or even general overall shifts in the market. One particularly
difficult prediction is foreign exchange rates between different currencies.
Because there are multiple players, trying to exploit small niches to make
money. They create the market and its behavior. Because of many people playing
with different information, the market is commonly thought to be “efficient”.
Here “efficient market hypothesis” is based on the premise that there is little
opportunity to exploit the market changes because as soon as a small one occurs
that might be predictable, someone probably already beat you to it. It also implies that the price of anything in
the market has been efficiently set to the correct value for the current time
on the basis of future risk and future possibilities for profit. Because of
this it is generally believed that it is difficult to predict future market
behavior based on historical information in any better manner than basically at
random. In the case presented here, 4 years of exchange rate data between the
Swiss franc and the U.S. dollar were obtained for the years1985-1988. A
standard feedforward neural network algorithm (no recurrent paths between nodes
of different layers) was used with the backpropogation learning algorithm. Two
network architectures were tried. The first had seven input units, seven hidden
units, and one output unit.; the second had an input layer of seven units, two
hidden layers (five units & two units) and a single output unit. Both networks individually, as well as the
average of their outputs were tried. This averaging seemed to improve the
performance of either network alone and was tried in an effort to remove any
bias inherent in using one particular architecture. The filtering function used
at each hidden node and output node took the summed inputs of the node and
output a value between -1.0 & 1.0.
The output to be predicted
was the direction of the exchange rate (1.0 for an upswing and -1.0 for a
downturn)
5. Case
Questions:
(i)
What are the benefits of using neural
network algorithm?
(ii)
How does the back propagation learning
algorithm works?
(iii)
Why the two hidden layers are used?
(iv)
Discuss the weakness of this model if
any?
No comments:
Post a Comment