Introducing myself - GSOC '15.

28 views
Skip to first unread message

shantanu prakash

unread,
Feb 24, 2015, 9:28:28 PM2/24/15
to raxa-jss-em...@googlegroups.com
Hello! 

I am a third year undergraduate at the Dept of CS of BITS Pilani, Pilani. I have had experience in Java, C/C++ and Ruby. Have a decent grasp on the CS concepts, particularly interested in ML and algorithms. I have worked with JS, RoR and Android. I wanted to contribute as a developer for the summers of 2015. 
Having had experience with Machine learning and some heuristic algorithms circulating artificial intelligence technologies, i thought i could contribute to the project: Intelligent scheduler.

Thank you!

shantanu prakash

unread,
Feb 27, 2015, 4:27:54 AM2/27/15
to raxa-jss-em...@googlegroups.com

Hello everyone!

I have setup OpenMRS on my system and have started familiarizing with it's architecture and functionalities. I have also done a Sencha touch FirstApp for initial exposure.
Since i have been interested for the project "Intelligent Scheduler ", i have been spending time reading up papers and studying other approaches online. 

What i figured was that since we are focused on making an an application which effectively schedules appointments for doctors, we could take two approaches:
1. This can be implemented using a future ailment prediction model for the patients (of the doctor). Here, using extensive data of medical records, for a patient’s sequence of symptoms we can try predict the rest of the sequence (and ailment), from other patient’s data. A bayesian probabilistic model can be used to implement this.
2. Another approach can be to study the patient's appointment pattern and his current diagnosis, and hence predicting his next visit. A logistic regression approach can be taken here.   

The papers I found, and the research that has been done is a lot on predicting ailments, rather than detecting pattern for visits. There is some correlation but a different approach is required, we might not want to be focusing on the giant "ailment prediction" problem, when we're trying to make a scheduler app.

BAYESIAN HIERARCHICAL RULE MODELING FOR PREDICTING MEDICAL CONDITIONS http://www.stat.columbia.edu/~madigan/PAPERS/rec_sys23_ssrn.pdf

Predicting A Patient's Future Medical Visits: A Comparison Of Quality Of Life And Clinical Variables http://www.valueinhealthjournal.com/article/S1098-3015%2813%2901381-8/abstract

Can seniors please guide on the approach i am taking!

Thank you!

Shantanu Prakash

Reply all
Reply to author
Forward
0 new messages