While going through machine learning topics, I got the point about classification and clustering.
I want to understand more about the business use of these definitions?
For example, on which two business needs specifically decide classification or clustering.
Many thanks in advance for an answer
Although there might be similarities, they have very different use cases. Clustering is used predominantly if companies want to ‘openly’ identify some common demographics within their customers, say you have sales data with customer information then clustering would yield info on groups of customers with similar purchase or demographics. Classification on the other hand is a bit more ‘targeted’ i.e they know what they are looking for. E.g based on historical purchase info they can identify which customers would make a specific purchase or cross a specific purchase value
Thanks, @sendilab I just want to confirm my understanding based on your above comment,
“You suggest clustering normally a good choice when business targeting unknown data, on the other hand, classification is a good choice when a business holds some background about data” - correct?
So clustering is basically a part where you look for patterns,structures in data thereby creating clusters & then if you start classifying /labelling those it turns to classifying.So to go for business perspective let’ s say business has entire past data but then to gain insights on what group of customers where interested with ofcourse their relevant data would help you to group them.Now when a new person jumps in as customer you basically decide which part would he be interested in which would be classifying.So the two are very related & thus there’s something termed as ‘Semi-Supervised Learning’ that is a merger of both the process.Hope I might have shed some light!!
Thanks, @yashrajtambe54 for your clear comment. helpful one