Workshop content

NBIS_NEUNETDL_V26 • Neural Networks and Deep Learning

List of presentations
Authors
Affiliation

A note on usage

Html slides have been authored using revealjs and can be viewed directly in the browser. Press s to view speaker notes.

Slides

Session Title Description
RNN Recurrent neural networks Introduction to recurrent neural networks
Basics Programming and training Neural Networks with PyTorch First NNs with PyTorch
Basics Programming and training Neural Networks with PyTorch, part 2 Second NNs with PyTorch
Good Practices Good practices of NN/DL project design DOs and DONTs when handling your data
CNN Convolutional Neural Networks for Image Classification Convolutions, pooling, BatchNorm, and CNN architectures (AlexNet, VGG16, ResNet) for image classification
CNN Convolutional Neural Networks for Image Segmentation Object localisation, detection (YOLO, RetinaNet, Faster R-CNN), and semantic segmentation (U-Net) with CNNs
No matching items

Labs

Session Title Description Jupyter Notebook
RNN Lab session: predicting time series Airplanes lab airplanes.ipynb
RNN Lab session: predicting time series, discrete state space Alphabet lab alphabet.ipynb
Basics Lab session: your first Neural Networks PyTorch intro exercises lab_introduction_to_pytorch.ipynb
Good Practices Target leakage example: classifying text data (tweets) Target Leakage lab investigating_target_leakage.ipynb
Good Practices Rigorous splitting of datasets into train and validation Datasets lab rigorous_train_validation_splitting.ipynb
CNN CNN Lab 1: Classification of Human Blood Cells using PyTorch Classify human blood cell images with PyTorch, progressing from a minimal CNN to an advanced CNN and finally to VGG16. lab_classification.ipynb
CNN CNN Lab 2: Segmentation of Human Blood Cells using PyTorch Pixel-wise classification of blood cell images with a U-Net in PyTorch, then the same task with a VGG16-backboned U-Net from segmentation_models_pytorch. lab_segmentation.ipynb
No matching items