Update README.md

Author: alik604

Intrusion-Detection/KDD cup '99/README.md

@@ -2,9 +2,9 @@
 
 > The KDD’99 an Intrusion Detection Dataset has 41 dimensions of data and a very large about of samples (“rows”), given that Sklearn is CPU only and that much of the code is a single-threaded, the training time is often unreasonable long. This is not a technological challenge
 
-+ `Intrusion Detection V2 99.99acc Multiclass_CLF.ipynb`: The best and most recent notebook
++ `Intrusion_Detection_Multiclas_V2.ipynb`: The best and most recent notebook
   + [Blog post](https://medium.com/@alik604/predicting-the-nsl-kdd-data-set-with-98-accuracy-240a7a245c9d)
 
-+ `Intrusion_Detection.ipynb`: Multiclass Classification
-+ `kddcup_99_100accAchieved.ipynb`: Singleclass Classification
++ `Intrusion_Detection_Multiclass.ipynb`: Multiclass Classification
++ `Intrusion_Detection_Singleclass.ipynb`: Singleclass Classification
 + `KDD_autoencoder.ipynb`: Ignore this, instead look at `UNSW_NB15 - PyTorch Categorical Variational AutoEncoder with Gumbel Softmax.ipynb` & `UNSW_NB15 - PyTorch MLP and autoEncoder.ipynb`, it is a much better example.