Please! help me understand these terms

Hi everyone, I know that maybe my question has nothing to do with the ones that this forum frequents, but the thing is that I have a job from University on terminology and by the terms they have given us, we are pretty sure that it’s about Machine Learning.

However, my colleagues and I lack so much knowledge on the subject and we find it difficult to relate the terms, with which we must develop an outline, these are the terms:

-Risk function *

-Optimal error *

-Composite outlier*



-Decentralized detection *

-Multitask linear *


Some of these terms we have been able to understand, sadly the ones with the asteriscs are the one that we have more trouble with.

We have been documenting ourselves for a two weeks but we still do not find articles or texts that clearly define these concepts, since most of what we find are graduate works, where they only focus on some parts of Machine Learning. And sadly, we do not have much time because on top of this, we are in our final exams…but we are really trying our best with this project.

We would appreciate it if you could give us some details about these terms or how they can be related to each other, even a little info about them would be awesome!

Thank you very much,


These are all statistics terms. Try a university-level guide to core statistics? I’m unsure why you’re doing machine learning stuff without having been taught what would seem to be critical maths/statistics modules beforehand but :man_shrugging:t3:

I know it seems weird…but this is what our professor gave us so :confused: we have to do terminology documentation work and relate these terms to make an outline
I’ll try to do more research but i dont really know where to look for it haha, all that i have found are dissertations about regression method or problems of regression…so i dont really know how to stick everything together…
Thank you for the information you have given me

I guess it makes sense to research terms for work/modules you’re going to do in future. But yeah basic and university-level statistics: current ML is mainly just application of statistics & you definitely shouldn’t need to dig through graduate-level stuff to get definitions.

well … the funny thing is that my area of study is not even related to mathematics or statistics or anything like that … and I think our teacher knows that we lack this type of knowledge, but yes, I will try to synthesize information and make it as simple as possible.