XGBoost notebook changes

1. removed the older im-xgboost directory
2. added xgboost_mnist directory within amazon algos for multi-class classification
3. added xgboost_abalone directory within amazon algo for regression
4. modified the README.md for amazon algos
5. added xgboost_bring_your_own_model to adavanced

Author: vrkhare

advanced_functionality/xgboost_bring_your_own_model/README.md

@@ -0,0 +1,2 @@
+### Bring Your Own Model (XGboost)
+`sagemaker-xgboost-bring-your-own-model.ipynb` shows how to train and Xgboost model in scikit-learn and then inject it into Amazon SageMaker's first party XGboost container for scoring. This addresses the use case where a customer has already trained their model outside of Amazon SageMaker, but wishes to host it for predictions within Amazon SageMaker.