• Stars
    star
    1
  • Language
    HTML
  • Created about 4 years ago
  • Updated about 3 years ago

Reviews

There are no reviews yet. Be the first to send feedback to the community and the maintainers!

Repository Details

More Repositories

1

Text-Moderation

The objective of text moderation model is to detect any text that contains curse words or inappropriate content (e.g., Hateful speech, violence). The model is based on BERT (Bidirectional Encoder Representations from Transformers) which is a Google AI Language research paper that introduces a bidirectional training of transformers that enable transfer learning for different NLP (Natural language processing) tasks. More Info can be found on pdf
Jupyter Notebook
2
star
2

Cheating-Detection

Jupyter Notebook
1
star
3

NEURAL-NETWORK-VS-CONV-NETWORK-FEATURE-SELCETION-BINNING

Importing Packages Reading Dataset Dataset Analysis and Prepartion Defining the binary classification column Identifying X and Y Sampling process Since we cannot use '?' when sampling we will convert it temporary to -1 Doing Sampling Process Defining Network and helper function Defining Neural Network Defining Convilutional Neural Network Defining Helper Function Using Neural Network Confusion Matrix Using Convloution Nueral Network Confusion Matix Comments Feature Selection Feature Importance Using top 5 Features Build Neural Network Create CNN Model Using Top 10 Features Creating Neural Network Creating CNN model Using top 15 Features Creating NN Model Creating the CNN Model Using Top 20 Features Creating Neural Network Creating CNN Model Using Top 25 Features Creating Neural Network Creating CNN Model Using all Features Creating Neural Network Evaluating Results Confusion Matrix Feature Binning and Discretization Using pandas qcut functionality to make the transform Feature Importance Using top 5 Features Build Neural Network Create CNN Model Using Top 10 Features Creating Neural Network Creating CNN model Using top 15 Features Creating NN Model Creating the CNN Model Using Top 20 Features Creating Neural Network Creating CNN Model Using Top 25 Features Creating Neural Network Creating CNN Model Using all Features Creating Neural Network Evaluating Results Confusion Matrix COMPARING WITH AND WITHOUT FEATURE BINNING Convolutional Neural Network Evalution Neural Network Evalution
Jupyter Notebook
1
star
4

advanced-encryption-standard

The objective of the following assignment is to implement a fully working AES algorithm that can encrypt or decrypt the given input. The AES cipher includes many stages and modules to implement. As assistance, I am going to list the main concepts you need implemented to have a working AES cipher. Note that the AES we are building uses a 16-byte key. ▪ ▪ ▪ ▪ ▪ ▪ ▪ S-Box – you need to implement how the byte substitution works in case of encryption and decryption. You will be given a file containing the tables needed. Row Shifting – Simple row rotations according to the AES rules. Mix Columns – This is the tricky part. This matrix multiplication in this operation is done in the field GF(2 8 ) which means that all arithmetic operations are different. So, for this part, you will first need to implement both addition which is XORing and multiplication which is a series of Shifting and XORing operations. (Look it up in your courseware) Add Round Key – Simple XORing between the state and the round key. Key Expansion – Follow the procedure for generating the 44 words required for the 11 keys (first key and one key for each round). AES Function – finally, you will implement the AES function itself where the magic happens. Later, you will implement the 10-round functionality. After finishing your AES encryption algorithm, tweak it so that it does decryption instead of encryption. Hint: Use the reversible technique. That is implement decryption in the same manner as encryption
C++
1
star