Chapter 9.9, Problem 9.5ARQ

Explanation of Solution

Data Mining:

Data Mining is the extraction of knowledge and data patterns from various raw data sets by examining patterns, trends and other Business Intelligence reports using intelligent methods for classification and prediction.

• Data mining techniques differ from reporting applications, as they are very sophisticated and complex, hence difficult to use.

Difference of factors for reporting and data mining:

Factors	Reporting	Data mining
Type of objective	Assessment	Prediction
Company	Target	Netflix
Analysis	Simple-summing, totaling	Advance statistics
Types	Noninteractive – RFM, Interactive - OLAP	Cluster Regression Market basket Decision tree Others

Artificial Intelligence and its potential benefits:

Artificial Intelligence (AI) is the ability of machines to perform activities that require human intelligence. In AI, machines can have vision, and can perform communication, recognition and learning. In AI, machines also have the ability to make decisions.

Benefits:

• Dealing with heavy and mundane tasks become easier with the help of machines.
• In order to gather and analyze Big Data, AI is extremely useful to improve efficiency.
• AI will potential increase cyber security and improve the security of Internet of Things (IOT).
• The accuracy of working on a thing increases a lot with AI.
• Using AI the use of digital assistants will increase which in turn will decrease the need for human resources.

Difference between Data Mining and Machine Learning:

Data Mining	Machine Learning
Data Mining is the extraction of knowledge and data patterns from various raw data sets by examining patterns, trends and other Business Intelligence reports using intelligent methods for classification and prediction	Machine Learning uses various data mining techniques to extract knowledge from data based on algorithms and is a subset of Artificial Intelligence.
In order to find patterns among data, Statistics and other programming methods are used in Data Mining to explain certain phenomenon without prior training.	Based on the previously known training data, one can predict the outcome using Machine learning.
Data Mining uses both Math and programming methods but inclination toward maths is more.	Machine Learning uses Data Mining techniques to build models that mostly use programming more than maths.

Data mining techniques are difficult to use:

Curse of Dimensionality:

The Curse of Dimensionality is the observation that is observed that problem arises when one analyses and organizes the data in high dimensional spaces. Working with data becomes more demanding with increase with increase in dimensions.

• With the increase in number of attributes, there is more chance to build easily a model to fit all the sample data but as a predictor it is useless.
• In data mining analyses, having too many attributes is problematic as one of the major activities in Data Mining concerns efficient and effective ways of selecting attributes.
• The amount of data used for Data Mining is huge and one needs to reduce the volume the data in order to meaningfully analyse the data.

Difference between Supervised and Unsupervised Data Mining:

Unsupervised Data Mining	Supervised Data Mining
In Unsupervised Data Mining, before running the analysis, analysts do not create a model or hypothesis.	In Supervised Data Mining, before running the analysis, data miners create a model and apply statistical techniques to the data.
Cluster analysis is a technique that uses Unsupervised Data Mining	Regression Analysis is a technique that uses Supervised Data Mining.

Cluster Analysis:

• Cluster Analysis is a way of arranging data such that data having similar properties are grouped together in a cluster. It is also known as clustering.

Example:

• Using Cluster Analysis, one can find patients with similar diseases from medicine history and demographic data.

Regression Analysis:

Data mining analysis which processes the consequence of a set of variables on other variables is called a regression analysis...