Introduction to Deep Learning

Build and train your first neural network in this practical two-day deep learning workshop for researchers, held in Amsterdam on 22–23 September 2026. Taking place from 09:30 to 17:00 at the Netherlands eScience Center, Science Park 402, this in-person course offers a structured route from traditional machine learning into deep learning. Through instructor-led coding and hands-on exercises, you will learn how neural networks work, implement models in Python with Keras and develop the skills to evaluate and improve their performance. What you’ll learn The workshop follows the Carpentries teaching style: instructors write code on screen while participants follow along. Across the two days, you will work through the main stages of a deep learning workflow, including: Identifying suitable problems for deep learning and understanding neurons, inputs, outputs and loss functions Preparing and exploring data with Pandas and Seaborn Using one-hot encoding for classification Building and training fully connected neural networks with Keras Interpreting loss curves and evaluating results with confusion matrices Creating regression models and measuring their performance Using training, validation and test splits correctly Recognising overfitting and applying normalisation, dropout and other preventative strategies Building convolutional neural networks for image datasets Adapting a state-of-the-art pre-trained network through transfer learning Organising deep learning projects and identifying practical next steps By the end, you should be able to train your first neural network, troubleshoot its training process and understand how to continue improving your model. See the detailed schedule and preparation information on the workshop website. Who should attend?… Register for this event