I basically want to build a system which will provide a student step by step guide or you can say a full route about what courses he/she should take currently or what examinations they need to clear to ultimately land up in universities of their choice in future.

I want to this project for the process of learning and i need some starting points about how should i go about doing this .


1 Answer 1


First of all, interesting question!

Most important thing to recommend you a starting point is to know what kind of data you have in hand. So my answer is regardless of that fact.

If you have a road-map of studies of students who got into University XXX then you can use model it with an HMM (Hidden Markov Model).

In this case the observations are the sequence of courses one passed and you want to calculate the probability of landing in desired university. Transition matrix is the probability of getting course A after getting course B. Emission matrix might be something like how successful they were in the course. Transition and emission matrices can be scored using facts like dropping a course or failing or the education track they exhibit in.

  • $\begingroup$ The point is i dont know what kind of data i should be looking for.The road map suggested by you could be one such structure , or i thought of going down the pre requisite tree of courses.Suppose university XXX need a student to have A,B,C things to be completed, then A needs A1,A2,A3 and B needs B1,B2 and C needs C1,C2,C3,C4 and it goes like that. So could you please help me with that and also how can i find such data sets. $\endgroup$ Jun 19, 2017 at 9:21
  • $\begingroup$ Dataset is very special. You need to search for it. But about the structure you proposed; it's not much of Machine Learning. It's just a tree structure as it's used by any school traditionally (in my bachelor uni we called it Study Chart and in masters Curriculum, both the same thing!). You might put weights on edges based on the performance of student but it's still not much learnable despite you have such tree and grades for many students and then you use it as training set. Then when a new student comes, based on the performance in first semester, you find its most similar tree to guide him $\endgroup$ Jun 20, 2017 at 8:09
  • $\begingroup$ For algebraic methods u need convert trees to a n-dimensional point in a metric space and do machine learning or you can learn from graph. But before all, "Fine The Data"! Once you have data you can describe it here and I would be glad to help you proceed. $\endgroup$ Jun 20, 2017 at 8:10
  • $\begingroup$ I am actually new to all this stuff and kind of need some things that i should be learning before starting off on any such project.Could you please help me with any such topics or resources i should be reading and may be then i could search for some data set . Thanks a lot anyways! $\endgroup$ Jun 20, 2017 at 12:59
  • $\begingroup$ You may start with having a look at recommender systems and how they function (chapter 9 here : mmds.org ). That might be one approach. The other one is getting urself familiar with HMM to see if it suites your goal or not (this one di.ubi.pt/~jpaulo/competence/tutorials/hmm-tutorial-1.pdf). Hope it helps :) $\endgroup$ Jun 21, 2017 at 8:14

Not the answer you're looking for? Browse other questions tagged or ask your own question.