Repository for Sylvie Rheault
Background The general objectives of my brainhack project is to familiarise myself with machine learning techniques, develop effective way of sharing my work with a team, practice python programming and planify organization of the data for my PhD work to come.
Specific learning objectives and deliverables/n During the brainhack school, I hope to achieve to following goals:
-
Starting from a fictive equation and data calculated with the equation, test machine learning tools to find the weight of parameters for the linear model. For example :
f(t,u,v,w,x,y,z) = 2.1t + 3.6u + 2.8v + 5w + 1.4x - 2.5y + 4*z
f()=cognition
t=systolic blood pressure/140 | u=diastolic blood pressure/90 | v=glycosylated Hb/7 | w=age/70 | x=LDL/2 | y=years of education/12 | z=VS/20 |
-
Redo with a non-linear equation.
-
Redo with introduction of variability in the model. (Precision on the mesures of parameters, errors, etc)
-
Try the automatic clustering on one model.
-
Program in python a small algorythm to read text files medical past history and medication in tables (the problem is they used lot of different words for the same disease, abrviations, etc.)
-
Planify the organisation of my data.
I don't know with who I want to work with but I have lot of interest in vascular impact on cognition and my final model will probably includ that parameter.