Assignment 3 A critique of a seminal DMKD paper This assignment is to produce, in a manner of a critical review, a 10-12 page paper which discusses the importance of a research paper published in the...

Sequential Mining Patterns
(
[Type the company name]
2018
Sequential Mining Patterns
[Type the document subtitle]
hp-u
[Type the company address]
)
Table of contents
1) Abstract
2) Introduction
3) Data Mining
$) Types of Data Mining
5) Sequential pattern mining
6) Importance of sequential pattern mining
7) Need of Sequential pattern mining
6) Sequential pattern mining approach
7) General approach for sequential pattern
8) Generalized sequential pattern
9) Conclusion
10) References
Mining Sequential Pattern
Abstract
This paper analyzes and presents the pattern and algorithm analysis of sequential mining. A database sequence is given where transaction time ordered each sequence that is a list of transaction. Each transaction is a set of items. With minimum support that specified by user, there is problem to discover all sequential patterns. Where, support of pattern defined as the data sequence’s number that contained all pattern. Here we discussed the sequential pattern’s example. Sequential Pattern is 5% of customer bought Ring world and Foundation in a transaction, in later transaction, followed by second foundation
Introduction
In Broad application, mining sequential pattern is a very important mining technique. In various domains like sales record analysis, marketing strategy, natural disaster, medical treatment, shopping sequences and DNA sequences found very useful. Frequent relevant pattern from the given sequences are discovered by it. Here sequence database having sequence have been provided. Transaction time ordered the each sequence is a list of transaction. The number of items is consisting by each transaction. There are various mining sequential pattern algorithms such as SPADE, GSP and PrefixSpan. These find the frequent relevant pattern from the sequences. The important attribute for each data set is timestamp in this algorithm. In the data mining process, it is important to give the information that is useful and more accurate. In this paper the detailed survey of sequential mining algorithms is presented(Agrawal,1993). This is done by step by step, by their used approaches, categorized these algorithms to solve the mining problem. As per various features and performance point of view, we have compared one to another.
Data Mining
The process of running powerful algorithm on data to extract or mine knowledge or information that is useful from large amount of data is known as Data mining .in data mining The information or patterns that are extracted from large repositories of information such as data warehouses,, relational database and In this xml repository etc .technique, the need of uncover important data pattern and perform data analysis(Agrawal,1995). In various domains, it is useful such as: fraud detection, decision support, business management and market analysis. To extract information from input sequence, many approaches have been proposed. The one of the most important method is sequential pattern mining. In given database, the problem of the presence of frequent sequence discovered has been solved by this mining algorithm. The data sequence defined as the database given to this algorithm is the time ordered set of sequence as input data. There is a number of transactions in each data sequence where a set of literals known as items, contained by each transaction. Note that in transactions, the order of items or item set does not matter. The data mining is the core processes of knowledge discovery in database (KDD). Usually in KDD, there are three processes. In first one is preprocessing which includes data integration, data selection, transformation and data cleaning. The data mining process is the main process of KDD. To produce hidden knowledge, different algorithms are applied in this process. The another process is post processing, according to the domain knowledge and user requirement this evaluates the mining result. If the result is satisfactory, the knowledge can be presented regarding the evaluation result. Otherwise until we don’t get the satisfactory result, we have to run all of that process again. As we know the whole data of database are not linked to mining task, from integrated resources, there is a process to select task related to data. The format that is ready to mined, data transform into that format. On data sources, various data mining techniques are applied. As a mining result, different knowledge comes out. Certain rules such as domain concepts or knowledge are evaluated those knowledge(Agrawal,1994). The final step is visualizing the results after getting knowledge. It also displayed as tables, decision tress, rules, charts, raw data, 3D graphics and data cubs. The result of data mining is more understandable and easy to be used by try to make this process.
Types of Data Mining
In data mining, there are two classes. That is prescriptive and descriptive mining. In data repository, to characterize or summarize general property of data is descriptive mining. In prescriptive mining, the predictions based on the historical data and perform inference on current data. Generally, in data mining techniques, there are various types are such as classification, association rules and clustering. Sequential pattern mining and web mining are also well researched on the basis of those techniques.
Association
The data mining technique that is best is association. In association, on the basis of relationship between particular item on the other item in the same transaction is discovered a pattern. For example, in market basket analysis, the association technique is used to identify what products purchased by customer frequently. On the basis of data business, to make more profit, to sell more products(Agrawal,1995).
Classification
It is classical technique of data mining which is on machine learning basis. The mathematical technique like neutral network, linear programming, decision trees and statistics are used by classification method. In classification, in set of data, each item into predefined set of groups and classes. In classification, how to classify data into groups can learn by software. For example the application where past records of employee who leave the company given and current employee who stay in company are applied the classification. Means two groups are formed. Then data mining software classified the employee into each group(Han,2000).
Clustering
The useful or meaningful cluster of objects that have same characteristics are made using automatic technique is data mining technique clustering. Classes are defined by clustering technique and put objects in that class. Here we take an example of library to make the concept clearer. A wide range of topics of books are available in library. Books have some kind of similarities keep in one shelf or one cluster by meaningful name by using clustering technique. Readers easily grab the book whatever they want, goes through the shelf instead of whole library.
The different types of techniques of...