Mining Decision Trees Theory Applications

Discuss About The Mining Decision Trees Theory Applications?

 

Introduction

In this twenty first century businesses are experiencing a wide access to data. There has been an increase in the memory along with the computing power to the machines, business data has become more and more attractive (Aggarwal, 2015). The concept of the ‘big data’ has been utilized almost in every field. There are a lot of researcher who are dedicated to finding useful information from this mass data which is regarded as mining (Aggarwal, 2015). The data mining is an enterprise method intended to explore large amounts of data to discover meaningful patterns in addition to rules. Over the decades, the data mining has gone through under different names for example the enterprise intelligence, predictive analysis, information discovery as well as predictive modeling (Braha, 2013). The process of data mining entails finding of the patterns in a lot more complex data sets, with the purpose of synthesizing the data as well as utilizing it to render the predictions in relation to the future (Braha, 2013). It also provide the business research workers to utilize the big data to ensure it is way quicker, along with improved capital investment as well as operating decisions all through the business . The data mining is a crucial tool with regards to the business environment as well as running it more effectively (Shmueli & Lichtendahl Jr, 2017). The process of the data mining has been aided by the computer as well as goals to search out and analyze a particular facts from the huge sets of the data.

Although the data mining technology has been improved greatly over the years, there are new challenges which still emerge in regards to specific data structure for example the high dimensionality, as well detecting the joint effect variables. To be able to discover on the previous patterns which were unknown to be able to predict after that there is need for organization to overcome on these challenges. In this research it aims to discuss various concepts in relation to the data mining. Some of the things which are discussed would be models related to the data mining, benefits of the data mining and drawbacks, ethical and unethical related to data mining, techniques of the data mining.

Model related to data mining

There are two kind of models of operation in relation to the data mining which enables one to discover data of interest for purpose of decision making (Fan & Bifet, 2013).

Verification model:

In this model it takes the inputs from the user along with testing the validity of it against the data (Larose, 2014). The emphasis of the use of this method usually lies with the user who is much responsible for the purpose of formulation of the hypothesis as well as issues of query in regards to the data to be able to affirm or even negate the hypothesis.

Discovery model:

In this model it usually differ in regards to the emphasis in which highlights that it is the system which automatically discover on the significance of the information which are hidden in the data (Fan & Bifet, 2013). The data is shifted in the search for the patterns, trends as well as generalization which occurs more frequently about the data without the intervention or even the guidance from the users.

Benefits of the data mining

Data mining entails collecting, processing, storing as well as analyzing the data to be able to discover any new information from it (Linden & Yarnold, 2016). There are many benefits which are associated to data mining they are as follows;

Marketing:

The data mining can help the marketing organization to be able to build on the models which is based on the historical data to be able to predict who would be respond to the new marketing campaigns for example the direct mail, the online as well as the marketing campaign (Rokach & Maimon, 2014). Through the result obtained it would be possible for the marketers to have an appropriate approach and be able to sell profitable products to their customers.

Finance/ banking

The concept of the data mining provides the financial organization such as the banks data in regards to the loan information and the credit reporting about the clients. They will be able to build models from the historical data of the customer and be able to determine the good as well as bad loan (Linden & Yarnold, 2016). Moreover, data mining could help the banks to be in a position to identify fraudulent credit card exchange to manage to safeguard the credit card owners.

Manufacturing

Through the application of the data mining especially in the operational engineering data, the manufactures are able to detect on any of the faulty equipment as well as determine on the optimal control aspects (Linden & Yarnold, 2016). An example, the semiconductor manufactures encounters the challenge that even the various conditions which are found in the manufacturing environment at different production plants are much similar (Linden & Yarnold, 2016). Data mining could be utilized to determine the ranges of controlling these parameters which lead to the production of the products at the desired quality.

Governments

The aspect of the data mining has helped various agencies of government to dig and analyze the records for the purpose of the financial transaction to build the patterns which are able to detect the money laundering or even the criminal activities.

Disadvantages of the data mining

The privacy concerns:

Data mining leads to issues relating to personal privacy which have increased over the years particularly when the internet has been booming with the social networks, forums, e-commerce as well as the blogs (Lu, Setiono & Liu, 2017). Due to the issue of privacy individuals are much afraid of their personal data being collected and utilized in a manner that is unethical which could cause them a lot of trouble. The business usually collect the information in regards to their clients in different ways to be able to understand on their purchasing behavior patterns.

Security concerns:

The aspect of security is a very big issue. The business possess data about their workers as well as the customers which include the social security numbers, as well as their payrolls (Rajola, 2013). There are concern on how this information is taken care of in those organization. There have been situations in which the hackers get access to the system in the organization and steal big data of the customers from these organization; example of organization which have encountered these are Ford Motor Credit and Sony (Rajola, 2013). Due to the huge number of the personal as well as financial data which is available there has been rise of the credit card and identity theft issues which has become a problem.

The misuse of the data/ inaccurate data

Data is collected by means of data mining which is intended for the ethical objective and in some cases this data could be misused (Rajola, 2013). The information may be subject to exploitation by unethical individuals or even businesses to be able to take advantage of the vulnerable individuals or discriminate against a given group of individuals.

Major Data mining techniques

There are several core techniques which are used for the data mining and they can make an organization to create an optimal results.

Classification Analysis

In this technique it is used to retrieve any vital as well as relevant information from a given set of data. This method is used to classify various data to various classes (Larose, 2014). The classification is much similar to the clustering in a way it could segment the data records into the various segments regarded as the classes.

Association rule learning

This refers to the technique to which can assist one in identifying some interest in relations to the various variables especially in the large databases (Larose, 2014). This technique could help one to unpack some of the hidden patterns in a particularly data which could be used in identifying of the variables that within a data as well as the concurrence of different kind of variables which appears more frequently in the datasets.

Clustering technique

This technique is a collection of the objects of the data and the object are much similar within a given cluster. This therefore means that the objects are significantly comparable to one within the same group and they differ to the objects in the other groups or to the other cluster. The clustering approach is the strategy of discovering the groups along with the clusters in the data in a way to which the degree of the association between the two objects is higher in case they fall under the same group.

Ethical and unethical issues related to data mining

There has been many arguments that the data mining is ethically neutral. Some of these arguments includes aspect such as how the data mining does not present new ethical issues, as well as privacy laws which are in place to be able to protect individuals and how the data mining is just another kind of statistical measures.

One of the ethical aspect is that the data mining is just like any other statistical procedure for example the surveys as well as the regression analysis (Witten, Frank, Hall & Pal, 2016).  Data mining only utilizes set of established as well as ethically accepted statistical approaches. Data mining is just a method of the data collection and interpretation.

There are privacy laws which are in place to be able to protect the consumers from any kind of harm. Moreover, many of the organization releases data about their privacy protection in respect to their consumer individual data (Witten, Frank, Hall & Pal, 2016). The law is therefore important when it comes to the ethics of the data mining. The privacy of personal data is an ethical issue and there is need for necessary laws which are in place to protect individual’s information from falling on the wrong hands which could cause a lot of trouble when the information is exposed.

Another ethical issue is the group profiling and this issue has been there even before the existence of the data warehouse as well as the data mining approaches (Witten, Frank, Hall & Pal, 2016). The data mining usually takes place when there has been a large collection of the data and this could present a new issue when it comes to the group profiling that has never occurred as before.

Unethical issues in data mining

There has been many arguments which have been present in regards to how data mining has been unethical practice (Rokach & Maimon, 2014). Some of these argument highlights on issues such as the loss of the privacy, the data not been utilized for the intended purposes as well as the issue of the group profiling. It has been argued data mining concept is unethical as it takes away on the rights of the clients over their private data.

Another issue is when the data mining approaches are done on the data which was collected for another purpose to which it was intended not to be. The ethical aspect is that the rights of the data are only released for a given purpose, and when they have been used for another purpose it could lead to adverse effect and yield negative consequences (Wu, Zhu, Wu & Ding, 2014). This can have unjust consequence to the person who had disclosed the information to the organization.

Another issue is that of group profiling especially when data mining takes place to scales which are much higher than the preceding techniques and this could led to the discrimination (Zhao, 2015). With the extent to which the data mining is performed on large data warehouse, the scales of the discrimination could be big. This could lead to the unethical discrimination against person who are based on the relationship which is highlighted in the two variables.

Conclusion

Data mining is a technique which has been utilized by various parties and it affect a large group of the stakeholders. The concept of data mining has become an interesting when it comes to the ethical topic and at such it has been analyzed to reach a balanced conclusion. From the research which has been done, it has highlighted on various concepts such as models related to the data mining, benefits of the data mining and drawbacks, ethical and unethical related to data mining, techniques of the data mining.

References

Aggarwal, C. C. (2015). Outlier analysis. In Data mining (pp. 237-263). Springer International  Publishing.

Braha, D. (Ed.). (2013). Data mining for design and manufacturing: methods and applications (Vol. 3). Springer Science & Business Media.

Fan, W., & Bifet, A. (2013). Mining big data: current status, and forecast to the future. ACM  sIGKDD Explorations Newsletter, 14(2), 1-5.

Freitas, A. A. (2013). Data mining and knowledge discovery with evolutionary algorithms.  Springer Science & Business Media.

Larose, D. T. (2014). Discovering knowledge in data: an introduction to data mining. John  Wiley & Sons.

Linden, A., & Yarnold, P. R. (2016). Using data mining techniques to characterize participation   in observational studies. Journal of evaluation in clinical practice, 22(6), 839-847.

Lu, H., Setiono, R., & Liu, H. (2017). Neurorule: A connectionist approach to data mining. arXiv  preprint arXiv:1701.01358.

Rajola, F. (2013). Data Mining Techniques. In Customer Relationship Management in the Financial Industry (pp. 109-125). Springer Berlin Heidelberg.

Rokach, L., & Maimon, O. (2014). Data mining with decision trees: theory and applications. World scientific.

Shmueli, G., & Lichtendahl Jr, K. C. (2017). Data Mining for Business Analytics: Concepts, Techniques, and Applications in R. John Wiley & Sons.

Witten, I. H., Frank, E., Hall, M. A., & Pal, C. J. (2016). Data Mining: Practical machine learning tools and techniques. Morgan Kaufmann.

Wu, X., Zhu, X., Wu, G. Q., & Ding, W. (2014). Data mining with big data. IEEE transactions on knowledge and data engineering, 26(1), 97-107.

Zhao, Y. (2015). Data mining techniques.