Medical datasets are classified using data mining and machine learning methods and techniques. How to categories various methods under data mining and machine learning.
These two concepts are closely related and sometimes share common ground in many applications. From my experience, machine learning involves the automated extraction of information using carefully authored algorithms and assigning AI patterns on a specific area of interest. On the other hand, data mining requires human intervention to apply machine learning methodologies on specific only datasets, in order to improve or deduct new classification patterns.
A similar question has been asked before in "RG Questions" area, where a couple of interesting and extensive answers were arised on machine learning versus data mining techniques. I've attached this post below my answer as well.
From Konstantinos Kontakis: "[...] data mining requires human intervention to apply machine learning methodologies on specific only datasets [...]"
I have heard this quite a few times, and it is not entirely accurate. Running data mining does not need a human to do anything other than to press 'go'. You may want a human to inspect the quality of the mined knowledge because data mining generally deals with subjective quality measures (e.g. 'interestingness'). Humans, with their domain knowledge, are best placed to quantify this.
As a concrete example, the fundamental algorithms for learning decision trees (e.g. ID3/CART) are situated very clearly in data mining pipeline (they require specification of, e.g., searching techniques, rule support, quality/purity measures), and you certainly do not need a human to inspect the splitting criteria. Decision trees, together with other tasks including rule set induction, are used within many fields, but they probably live most accurately in predictive data mining.
Data miners also work on descriptive data mining, in which sub-fields like association rule learning, subgroup discovery and relational data mining live. Such descriptive models may be able to discover and deliver knowledge to practitioners, and this knowledge may actually increase the understanding about the application. Perhaps this is what you meant, but to obtain these rules/knowledge we did not need a human to operate directly in the rule discovery algorithm.
What you say might be true if you adopt a data mining strategy like CRISP-DM, which defines an iterative process (possibly with human evaluators) for discovery of knowledge from databases.
Data mining is the process to discover knowledge or previously unknown pattern from structured or unstructured data. To do that, machine learning algorithms can be used.
On the other hand, machine learning is the study, design and development of algorithms that makes a computer to learn.
In case of categories methods under data mining and machine learning, it's little bit ambiguous in my opinion as both field closely related with each other.
Data mining and machine learning used to be two cousins. They have different parents. Now they grow increasingly like each other, almost like twins.
The field of machine learning grew out of the effort of building artificial intelligence. Its major concern is making a machine learn and adapt to new information.
The field of data mining grows out of knowledge or hidden pattern discovery from databases.
These two paradigms are used to analyze data but as far as I know, people of machine learning do not like the term of data mining to explain the process approached for machine leaning techniques.
I recommend you this link on the web that I have used for my students: