Week 4 assignment
- Define data mining. Why are there many names and definitions for data mining?
A process through which unknown data patterns in a dataset are discovered (Hand, 2007). Another description may be a method that uses mathematical and computational artificial learning techniques to derive useful data from the data and provide helpful information for large datasets. Data mining covers any form of automated data processing. Data mining can also be described as the technique for discovering similarities within the dataset. Such patterns can be models, averages, associations, affinities, or laws of prediction. Many definitions of data mining are because of vendors who have stretched the limits to include more forms of information analysis and increase sales.
- What are the main reasons for the recent popularity of data mining?
The most obvious reasons for the increased popularity of data mining include:
- The movement toward the de-massification of business practices.
- The significant drop in the cost of software and hardware for information processing and storage.
- Increased storage technology and processing of data
- The consolidation of repositories and databases into a single data warehouse.
- Integration and consolidation of records enable a view of vendor, transactions, and customer in a unique look.
- The recognition of the value hidden in large datasets
- Competition from a global level which is driven by changing consumer behaviors
- Discuss what an organization should consider before deciding to purchase data mining software.
It is necessary to find standard criteria to use in significant software before a company can decide to buy data mining software. Therefore, a look at the comparison of costs and benefits is essential. Often, look at the people’s skills who will be interacting with the program and perform the research. Besides, look at the availability of the data needed for data mining to be performed.
- Distinguish data mining from other analytical tools and techniques.
Data mining is among Data Processing practices. Data Analysis is a complete set of activities responsible for collecting, preparing, and modeling data to extract meaningful insights or knowledge. These are also used in the Business Intelligence sub-set. Studies on data mining are all for organized data. Data analysis on both structured, semi-structured, and unstructured data can be performed. Data Mining requires no preconceived theory to establish a pattern or trend in the results. Data analytics, on the other hand, checks a given approach.
- Discuss the leading data mining methods. What are the fundamental differences among them?
The primary data mining methods (Amatriain, Jaimes, Oliver & Pujol, 2011).
Association Learning: mostly used to examine customer behavior; it establishes the correlation between variables in large databases.
Regression Analysis: data is analyzed to predict the future trend and relationships
Classification Analysis: relevant and essential information about data and metadata is collected and classified into different classes.
Clustering Analysis: in this data mining process, data is collected and related and unrelated data grouped in different groups.
Anomaly Learning: different data from regular data is observed. This method is mostly used in fraud detection and intrusion detection.
Differences between data mining methods
The relationship between information analyzed in regression analysis and the correlation between the independent and dependent variables is analyzed to predict future relationships. However, correlation is analyzed in association learning to find the hidden patterns within a dataset. Similarities and associations within the data are explored in Anomaly learning, but the emphasis is on recognizing the particular item of data.
In classification analysis, the researcher has the knowledge about the various groups in which the evaluation is done, unlike cluster analysis, where the research is conducted only based on information similarity.
References
Amatriain, X., Jaimes, A., Oliver, N., & Pujol, J. M. (2011). Data mining methods for recommender systems. In Recommender systems, handbook (pp. 39-71). Springer, Boston, MA.
Hand, D. J. (2007). Principles of data mining. Drug safety, 30(7), 621-622.