Pfthb

Currently, I develop a model using xgboost which accuracy is 92% and now I am trying to see the bias and variance of my model by plotting the learning curve.

Here is my code:

xgb_params <- list("objective" = "reg:linear",
 "eval_metric"="rmse",
 "eta"=0.05,
 "max_depth"=2

 )
watchlist <- list(train = train_matrix,test = test_matrix)

bst_model <- xgb.train(params = xgb_params,
 data = train_matrix,
 nrounds = 500,
 watchlist=watchlist,
 gamma=0
 )
e <- data.frame(bst_model$evaluation_log)
plot(e$iter, e$train_rmse, col = 'blue')

**The train and test error output is** 

[491] train-rmse:275.988190 test-rmse:285.262756 
[492] train-rmse:275.954712 test-rmse:285.229706 
[493] train-rmse:275.933258 test-rmse:285.215637 
[494] train-rmse:275.917206 test-rmse:285.209808 
[495] train-rmse:275.909515 test-rmse:285.203552 
[496] train-rmse:275.861633 test-rmse:285.165009 
[497] train-rmse:275.828766 test-rmse:285.123657 
[498] train-rmse:275.801086 test-rmse:285.097107 
[499] train-rmse:275.681793 test-rmse:285.020081 
[500] train-rmse:275.655884 test-rmse:284.991364 

And the curve is enter image description here

By looking at the curve, can anyone tell me if this curve is overfitted or not?

Now to plot the learning curve (where X-exis= observation size and Y-axis=error count) don't find any function or library exist in through which I can plot the learning curve easily.

So, can anybody help me on this topic?