Why look at case studies

Hi Sir,

What does it mean the word case study refers to ?

Here is my understanding, case study refers to the application built using AlexNet, Resnet, Inception network. so Is case study the word refers to real time application build things ?

Or case studies means classic network /

Case studies just means it is an example of a method.

1 Like

Hi Tom,

So here case studies refers to ALexNet, LetNet 5, Resnet since these are methods act an example for Convolution Neural Network ? from which we can derive ideas to build effective ConVNEt ?

or case studies refers the real time applications built over using AlexNEt , LetNet 5 Resnet ?

Can u please elaborate case study in the context of Letnet 5, ALexnet please ?

I think you’re making this more complicated than it needs to be.

Yes, looking at examples of applying LeNet or AlexNet would also be “cases”. Whether they are examples of real time applications just depends. Some of them are and some of them aren’t. The reason for looking at other examples of applying ConvNets or networks in general to solve a problem is that you’re hoping to gain insight about how to approach a particular problem that you want to solve, right? So you look for examples that are similar in as many ways as possible to the problem you have in front of you. Is it a real time problem or not? That will determine what kind of other examples of deep neural network applications would be most useful for you to examine and consider as “case studies”.

Sir , According to your points, can we conclude like, case studies refers to AlexNet, LeNEt5, Resent because these are considering as examples useful for building efficient convolutional Neural Network ?

According to Tom Mosher definition Case studies just means it is an example of a method, can we say Method is CNN and example is classic Networks (Alexnet, resnet etc)

Am I right sir ? or would like to correct anything my understanding ?

The horse is dead! Long live the horse! :laughing:

Sorry, I guess that for non-native speakers of English, that joke may not make any sense. Let me explain:

It’s a layered thing. There is an expression in English “to beat a dead horse” as in “I think we are beating a dead horse here …”. It is used when an argument or discussion has basically been settled, but people insist on continuing to do go over the argument again. If the horse is dead (the argument is settled), then what it the point of continuing to beat it.

So the first level of the joke is that I’m saying that I think you are “beating a dead horse” here.

But then the second level of the joke is that in England, when the King (or Queen) dies, the new King or Queen is installed essentially instantly. So the newspaper vendors on the street corners will shout “The King is Dead! The King is Dead!” (referring of course to the previous King) in an effort to sell more newspapers, but immediately everyone will say “Long live the King!” referring of course to wishing long life to the new King who has just been coronated.

So we are wishing long life to the horse that we thought was dead, but you have now revived to beat some more!

“The horse is dead! Long live the horse!”

SIr :smiley: :grinning_face_with_smiling_eyes: Good Joke…

Sir I dont want to beat ur statmenet…I would like to make the conclusion of my understanding aligned with your thoughts , please help me sir. To make sure Im on your page.

Sir , According to your points, can we conclude like, case studies refers to AlexNet, LeNEt5, Resent because these are considering as examples useful for building efficient convolutional Neural Network ?

According to Tom Mosher definition Case studies just means it is an example of a method, can we say Method is CNN and example is classic Networks (Alexnet, resnet etc)

Am I right sir ? or would like to correct anything my understanding ?

I don’t think I can help you any further here. Both Tom and I have made the same point a couple of times. A “case study” is a generic term that means looking at previous examples of some type of solution with the hope of gaining insight about how to apply those same techniques to other problems. It can be CNNs or generic DNNs or anything else. You are trying to turn it into some incredibly specific thing. What is the point?

Seriously, have you considered whether applying to Law School might be a better application of your talents? In Law School, they also do lots of “case studies”.

You are reading more detail into the concept than is appropriate.
A case study is just an example of a particular topic for discussion.

1 Like

I think the term comes from the medical field, where a person who is injured or sick is termed to be a “case”.

If a person goes to the doctor with a sore thumb, the doctor may study his situation (his case), and determine that he hit himself with a hammer while building a house.

Another case might be a person who has a broken arm, because they fell off a ladder.

Actually a well defined term, which context dependent on the topic area. Especially in a case of case competitions.

But from this thread it looks like in AI/ML there is no more value in “case study” that in “example” or “report”.