1. Definition of the concept
Machine Learning : This is broadly defined as “using experience to improve the performance of a computer system.” In fact, since “experience” in a computer system is mostly in the form of data, machine learning should try to analyze the data. Gradually, it has developed into one of the innovative sources of data mining technology, and in this regard, it is receiving more and more attention. To
Data collection : One explanation is “the non-trivial process of identifying effective, new, potentially useful, and ultimately understandable patterns in massive amounts of data.” As the name suggests, data mining is an attempt to find useful knowledge from huge amounts of data.
Two, relationships and differences
Relationships: Data mining can be seen as the intersection of database technology and machine learning. It uses database technology to manage massive amounts of data and also uses machine learning and statistical analysis to analyze data. The relationship is as follows:
Many disciplines have influenced data mining, including Database, Machine Learning, Statistics Undoubtedly the biggest impact. Roughly speaking, databases provide techniques for manipulating data, while machine learning and statistics provide techniques for analyzing data. Since the statistical community is often obsessed with the beauty of theory and ignores real utility, many of the methods provided by the statistical community usually need further study in the machine learning community and become effective machine learning algorithms before moving on to the field of data mining. In this sense, statistics mainly affect data mining through machine learning, and machine learning and database are the two supporting data mining technologies. To
Differences: Data mining is not just a machine learning application in industry. There are at least two important differences between them:
1. Traditional research in the field of machine learning does not consider large amounts of data as objects of processing, therefore, data mining must perform special, rather than simple, transformations of these technologies and algorithms.
2. As an independent discipline, data mining also has its own unique features, namely: Correlation analysis. Simply put, associative analysis is about finding out from the data that “people who buy diapers are more likely to buy beer” seems strange, but it might make sense.
image credit: https://www.foreseemed.com/