Semester page for FYS5429 - Spring 2026

Dear all, welcome back to FYS5429/9429. We hope the week has started the best possible way for you all.

Last week we wrapped up our discussions of recurrent neural networks (RNNs). Together with standard feed forward neural networks (FFNNs) and (CCNs) convolutional neural networks (covered the first 7 weeks of this course and for several of you a considerable repetition from earlier courses), CCNs, RNNs and FFNNs form the basic building blocks of more advanced deep learning methods. With these ingredients, we are now ready to start with entirely new topics. On Thursday this week we will start with Autoencoders (AEs) with applications to both supervised and unsupervised problems. On this Thursday's lecture we will discuss their mathematics and link (so-called linear AEs) with the PCA=principal component analysis.

AEs will us also serve us in linking (dimensionality reduction) to generative methods like Variational Autoencoders and Diffusion models, to b...

Dear all welcome back to FYS5429/9429.

The plan this week is to finish our discussion of recurrent neural networks, with a focus on the mathematics and how to write an RNN code and long-short-term-memory RNNs. After this (next week) we will discuss the whys and hows of autoencoders, before w emake a quick dive into transformers and then start with generative methods.

Plans for the week of March 2-6

Reminder on basics of recurrent neural networks (RNNs)

Mathematics of RNNs

L(ong)S(hort)T(erm)M(emory)s and RNNs

Writing our own codes for RNNs

Lab session: discussion and work on projects

Reading recommendations:

a. Goodfellow, Bengio and Courville's chapter 10 from?Deep Learning

b.?Sebastian Rashcka et al, chapter 15, Machine learning with Sickit-Learn and PyTorch

c.?David Foster, Generative Deep Learning with TensorFlow, see chapter 5

Th...

Welcome back to a new week with FYS5429/9429.

The aim of this week is to discuss recurrent neural networks (RNNs) and their basic mathematical operations. Our focus on applications will be on differential equations, however these data sets can easily be changed with those of your interest.? The lecture notes with code examples using PyTorch and Tensorflow/Keras are at https://github.com/CompPhysics/AdvancedMachineLearning/blob/main/doc/pub/week6/ipynb/week6.ipynb

Next week we will discuss long-short-term memory RNNs, and how we can write our own code for an simple RNNs, together with the basic mathematics needed. After that we will start discussing Autoencoders and Transformers, before we dive into generative models.

Else, at the lab we will discuss the various projects. The lecture is in-person but if you cannot attend in person, you can either join via zoom or watch the recording afterwards.

Reading recommendations

Dear all, welcome back to FYS5429/9429.

The plan this week is to end the discussions we started last week on CNNs, with a discussion on how to write codes (either our own of interest for those of you who would like to do so or with Pytorch and/or Tensorflow/Keras). Thereafter we will look at recurrent neural networks (RNNs).

The RNN part will be followed up by the autoencoders before we move into transformers and then generative methods.?

Note that this week's lecture is via zoom (I am away for a meeting). The lecture will also be recorded for those of you who cannot be there during the lecture. Oda and Ruben will be at the lab in person from 1215pm to 2pm (our normal times) and I will be available for the whole lab session via zoom if you wish to discuss with me.? The zoom link is the same as always, https://uio.zoom.us/my/mortenhj

Plans for the week of February 16-20

Implementing Convolutional Neural Networks (CNNs...

Dear all, welcome back to FYS5429/9429. This week the plan is to exclusively focus on how we can construct a code for convolutional neural networks (CNNs) and discuss the basic mathematics of the CNNs. We will continue with this topic next week as well, before we start with recurrent neural networks (RNNs).? For many of you these topics may be familiar, but together with neural networks and Autoencoders they define to a large extent what we may call the standard discriminative deep learning methods. We have thus decided to discuss these methods with an emphasis on the mathematics and how to write your own codes for either a CNN or an RNN (this is a possible project option, the so-called coding option).

After our discussions of these discriminative methods, we will move into a discussion of generative methods (mid March). These methods will keep us busy till the end of April, before we attempt an excursion into reinforcement learning (May).

The slides f...

Dear all and welcome back!??

Here are the plans for our sessions on February 5.? Note that since Morten is away for a workshop on quantum machine learning (Sweden), the lecture will be in the form of a recording only. However, we will run the exercise session 1215-14 as normal.? Thus, no regular lecture, just a recording. This will be uploaded before the regular lecture on February 5.

The teaching material is at https://github.com/CompPhysics/AdvancedMachineLearning/tree/main/doc/pub/week3/ipynb, see both jupyter-notebooks, the one called week3.ipynb and the one called BlackScholesPINN.ipynb.

Our plans for this week are?

Neural Networks with codes and Physics Informed Neural Networks, theory and codes

Start discussion of Convolutional Neural Networks (CNNs)

Discussion of possible projects during the lab session. We continue with your presentations during the lab session, please feel free to present your p...

Dear all,

We hope your week has started well. Below we outline the plans for the lecture on January 29.

The main objective of this week¡¯s lecture is to review the fundamentals of neural networks and to connect these concepts to the numerical solution of differential equations, with a particular focus on Physics-Informed Neural Networks (PINNs). While many of you may already be familiar with parts of this material, we believe it is valuable to revisit the core ideas, as neural networks form the backbone of many of the methods we will cover later in the course, including convolutional neural networks (CNNs), recurrent neural networks (RNNs), autoencoders, and related architectures.

During the lab session, we will also introduce and discuss different project variants. If you already have a preliminary idea for a project, you are encouraged to prepare 3¨C5 slides outlining your proposed topic. These can be uploaded to the course GitHub repositor...

Dear all, welcome to FYS5429/9429. Our first lecture is Thursday January 22 at 1015am to 12pm. We have also set aside an eventual lab session from 12.15pm to 2pm on Thursdays. Our lecture room is F?434 at the Department of Physics.?

You can attend remotely via zoom at the link https://uio.zoom.us/my/mortenhj

All lectures will be recorded and posted, together with whiteboard notes.

The emphasis is on deep learning algorithms, starting with the mathematics of neural networks (NNs), moving on to convolutional NNs (CNNs) and recurrent NNs (RNNs), autoencoders, graph neural networks and other dimensionality reduction methods to finally discuss generative methods. These will include Boltzmann machines, variational autoencoders, generalized adversarial networks, diffusion methods, reinforcement learning and more. ?See the course GitHub link for more information, weekly plans and more?https://gith...