No guarantee that the story is true. I used to do a particular type of modeling to help find potential oil fields. One thing I would do is deliver both an optimistic model and a pessimistic model, both of which would fit all the available data. Then it was management's problem.
“Data mining” as a concept has luckily faded away from general usage now due to how ineffective it was." - is this actually true? I am a beginner, just took a few courses here and there, I have thought (for 1-2 years I've been learning) that 'data mining' is HOT!
Relatively speaking there was a period in time when data mining was a hot topic w/ textbooks and articles being written about it (google trends seems to suggest this was in the early 2000s during the tail end of the dot com boom). Since then talk about data analysis, data science, and machine learning have become more popular than the term "mining" itself.
In a way you can consider the act of "Take a giant dataset and run unsupervised techniques to surface stuff" has been rebranded to certain kinds of ML. But the original label and the semi-manual methods of the past are definitely out of favor.
No guarantee that the story is true. I used to do a particular type of modeling to help find potential oil fields. One thing I would do is deliver both an optimistic model and a pessimistic model, both of which would fit all the available data. Then it was management's problem.
Step 1 is super important.
“Data mining” as a concept has luckily faded away from general usage now due to how ineffective it was." - is this actually true? I am a beginner, just took a few courses here and there, I have thought (for 1-2 years I've been learning) that 'data mining' is HOT!
Relatively speaking there was a period in time when data mining was a hot topic w/ textbooks and articles being written about it (google trends seems to suggest this was in the early 2000s during the tail end of the dot com boom). Since then talk about data analysis, data science, and machine learning have become more popular than the term "mining" itself.
In a way you can consider the act of "Take a giant dataset and run unsupervised techniques to surface stuff" has been rebranded to certain kinds of ML. But the original label and the semi-manual methods of the past are definitely out of favor.