Predicción del éxito académico en la carrera de Ciencias de la Computación de la universidad “Agostinho Neto”

Abstract

Since  its  creation  Computer  Science  course  of  the  University  Agostinho  Neto (UAN)  exhibits  a  very  low  academic  efficiency  (smaller  to  15%  on  average),  in
spite  of  the  urgent  necessity  of  having  qualified  personnel  in  Angola.  Therefore the  academic and  educational  authorities  are  concerned  to  improve the  process of teaching learning, especially being focused on the first year students. On this paper,  the  results  of  an  analysis  of  the  factors  that  affect  significantly  the academic  success  (in  the  context  of  the  UAN)  of  the  first  year  students  of  the Computer  Science  course  in  the  subjects  Logic  of  Programming,  Imperative Programming.  From the psychopedagogical and social point of view the reasons to select a group of factors are analyzed and the results are compared by mean of  a correlacional  analysis  for  estimating  the  possible  association  between  each independent  variable  and  the  variable  criteria.  Considering  above  results  a  few models that use the most outstanding variables to predict the academic success
in  each  case  are  proposed,  as  instances  of  a  more  general  model  based  on neural networks.

Key Words
Academic success, factors, correlation, neural networks.