baf-baf/deep_learning_lectures

deep_learning_lectures

This repository contains a miscellaneous collection of lectures and lab lectures on deep learning prepared for teaching activity during my PhD @ University of Modena and Reggio Emilia.

---

2017 - Lectures on Deep Learning @ UniMoRe, Italy

Crash course on Deep Learning and Temporal Data Processing.

THEORY

• 0_gradient_descent.pdf - slides on gradient descent optimization.
• 1_deep_neural_networks.pdf - slides on neural network and deep neural networks.
• 2_conv_neural_networks.pdf - slides on convolutional neural networks.
• 3_recurrent_neural_networks.pdf - slides on recurrent neural networks.

PRACTICE

SLIDES

• tensorflow_basics.pdf - slides on TensorFlow basics.
• linear_regression.pdf - Implementing a linear regression model in TensorFlow.
• neural_network.pdf - Implementing a fully connected network to classify MNIST digits.
• convnets.pdf - Implementing a convolutional network for classification on MNIST and segmentation on TilesDataset.
• lstm.pdf - Implementing a LSTM to count ones in binary sequences.

CODE

• lab00 - linear_regression
• lab01 - neural_networks
• lab02 - convnets for classification on MNIST
• lab03 - convnets for segmentation on TilesDataset
• lab03 - lstm_sequence_counting

LaTeX source

• Whole LaTeX source is available in repository in 2017_MASTER

---

2016 - Lab lectures, @ Machine Learning and Pattern Recognition course, UniMoRe, Italy

• In repo 2016_MLPR are contained lab lectures convering DNN, CNN, LSTM in Keras + Theano framework.

  Detailed syllabus can be found here.