Rede Neural Convolucional para reconhecimento de gestos em LIBRAS (Alfabeto) Projeto 01/2019 - Ciência da Computação (Universidade Anhembi Morumbi)

The project aimed to implement Deep NN / RNN based solution in order to develop flexible methods that are able to adaptively fillin, backfill, and predict time-series using a large number of heterogeneous training datasets.

Convolutional autoencoder for encoding/decoding RGB images in TensorFlow with high compression ratio

Avoiding the vanishing gradients problem by adding random noise and batch normalization

This package is a Tensorflow2/Keras implementation for Graph Attention Network embeddings and also provides a Trainable layer for Multihead Graph Attention.

Generic L-layer 'straight in Python' fully connected Neural Network implementation using numpy.

This project classifies SMS messages as spam or ham using a feedforward neural network in PyTorch with a bag-of-words representation. It includes train/validation/test splits, performance evaluation (accuracy, sensitivity, specificity, precision), and saving the trained model and vectorizer for reuse in inference.

Sentiment Classifier using Neural Networks & PyTorch.

PyTorch implementation of normalization-free LLMs investigating entropic behavior to find desirable activation functions