education-impact

A ML model to find which factors influence the development of the country the most

Approach

Initially we tried to predict country's development factors 40 years ahead, but due to data unavailability ,it's not feasible approach

Now we are trying to predict development factors 5-10 years ahead and we are considering only education levels of 20-25 age group people, reason for this approach i
in general most of the population education level reaches saturation point by the time they reach 25 and from 25 they start to contribute to the society ( working )
in this way by predicting 5-10 yrs ahead we can check how education is impacting the country's development

Random Forest

First Attempt

We trained a TensorFlow RandomForest without any data interpolation

train : test = 70 : 30

total data points : 2834

prediction years = 40

model performance:

"malnutrition","maternal_mortality","people_in_poverty" were not used by the model.

MSE : 5.545248985290527

RMSE: 2.354835235274546

Second Attempt

We trained a TensorFlow RandomForest with WCDE datasets, data interpolation

train : test = 70 : 30

total data points : 2834

prediction years = 40

model performance:

"malnutrition","maternal_mortality","people_in_poverty" were not used by the model.

MSE : 3.8452484607696533

RMSE: 1.9609305089088835