Merge pra mim

.gitignore

@@ -0,0 +1,16 @@
+bloco7_autoenconders/data/MNIST/raw/t10k-images-idx3-ubyte
+bloco7_autoenconders/data/MNIST/raw/t10k-images-idx3-ubyte.gz
+bloco7_autoenconders/data/MNIST/raw/t10k-labels-idx1-ubyte
+bloco7_autoenconders/data/MNIST/raw/t10k-labels-idx1-ubyte.gz
+bloco7_autoenconders/data/MNIST/raw/train-images-idx3-ubyte
+bloco7_autoenconders/data/MNIST/raw/train-images-idx3-ubyte.gz
+bloco7_autoenconders/data/MNIST/raw/train-labels-idx1-ubyte
+bloco7_autoenconders/data/MNIST/raw/train-labels-idx1-ubyte.gz
+bloco6_autoenconders/data/MNIST/raw/t10k-images-idx3-ubyte
+bloco6_autoenconders/data/MNIST/raw/t10k-images-idx3-ubyte.gz
+bloco6_autoenconders/data/MNIST/raw/t10k-labels-idx1-ubyte
+bloco6_autoenconders/data/MNIST/raw/t10k-labels-idx1-ubyte.gz
+bloco6_autoenconders/data/MNIST/raw/train-images-idx3-ubyte
+bloco6_autoenconders/data/MNIST/raw/train-images-idx3-ubyte.gz
+bloco6_autoenconders/data/MNIST/raw/train-labels-idx1-ubyte
+bloco6_autoenconders/data/MNIST/raw/train-labels-idx1-ubyte.gz

README.md

@@ -1,2 +1,26 @@
-# RedesNeurais
-Disciplina de Redes Neurais SCC0270
+# Redes Neurais e Aprendizado Profundo / Neural Networks and Deep Learning (2022)
+
+Compilado de programas, anotações e atividades da disciplina de SCC0270 - Redes Neurais e Aprendizado Profundo oferecida pelo ICMC-USP. Todos os programas desenvolvidos na disciplina foram feitos em linguagem Python com o framework Pytorch e possuem apenas fins educacionais.
+
+Compiled from programs, notes and activities from the subject SCC0270 - Neural Networks and Deep Learning offered by ICMC-USP. All programs developed in the course were written in Python with the Pytorch framework and are for educational purposes only.
+
+## A disciplina/ The discipline
+
+### Objetivos / Goals
+
+Apresentar ao aluno os conceitos básicos de Redes Neurais Artificiais e os principais modelos existentes. Analisar o comportamento destes modelos, suas capacidades fundamentais e limitações, possibilitando a utilização destas técnicas na resolução de problemas práticos.
+
+To present to the students the basic concepts of Artificial Neural Networks and the current most important models. To analyze the behavior of these models, their fundamental capabilities and limitations, allowing the use of these techniques to solve practical problems.
+
+### Programa resumido / Summary program
+
+Definição de modelos conexionistas. 
+Aprendizado em modelos conexionistas: aprendizado supervisionado, não-supervisionado, competitivo. 
+Arquiteturas básicas: Perceptron, Adaline, Perceptron Multi-Camadas, Redes RBF. 
+Aprendizado profundo: arquiteturas convolucionais (CNN), encoder-decoder, redes adversárias, transfer learning, redes recorrentes e modelos de atenção. 
+Sistemas de auto-organização: PCA, LDA e rede de Kohonen. Memórias Associativas: Redes de Hopfield. Aplicações.
+
+Definition of connectionist models. 
+Learning in connectionist models: supervised, unsupervised, and competitive learning. 
+Basic architectures: Perceptron, Adaline, Multi-Layer Perceptron, RBF Networks. Deep learning: convolutional architectures (CNN),encoder-decoder, adversarial networks, transfer learning, recurrent networks and attention models. 
+Self-organization systems: PCA, LDA and Kohonen network. Associative Memories: Hopfield Networks. Applications.