I am working with machine learning to research the retention of students using data from the past 7 years. Using predictive analysis we could intervene with the "most likely to leave" students and assist them with their struggles. The end goal is to increase the retention rate for programs in higher Education programs.
think Of Amblyopia if one is never used a eye it will go into blindness
so also if some one does not use the acquired knowledge it will slowly evaporate rate of evaporation or loosing this knowledge depends on the individul. When the student revisits the conept that is not retained, student can recall and make connections.
This is the basic difference between human and humanoid. Forgetting is essential to preserve mental health.
This is purely my experience working with children for over 20 years
I have not done this kind of analysis myself, but I think that many schools DO analyze trends to determine "students at risk." In this case, "risk" does not mean physical harm, but rather at risk of failure, and thus needing special support.
So, you are willing to construct the description of a student at a "risk", and extract this from your data. This is a normal task for data mining. This is realistic, you can build quite accurate model and have a good prediction.
There is definitely research out there. I cannot speak to their results, but, on a large scale, companies like CIVITAS Learning do this regularly. They develop institution-specific models.
The institutional research team at LaGuardia has done quite a bit of predictive modeling of retention. All papers are at http://laguardia.edu/IR/IR-facts/
The "Research Summary" summarizes much of the work and has links to the papers.
- Badges
- Science topic