Thesis offers

Boltzmann Machines are probabilistic models developed in 1985 by
D.H. Ackley, G.E. Hinton and T.J. Sejnowski. In 2006, Restricted
Boltzmann Machines (RBMs) were used in the pre-training step of
several successful deep learning models, leading to a new renaissance
of neural networks and artificial intelligence.

In spite of their nice mathematical formulation, there are a number of
issues that are hard to compute:

1. The computation of the partition function is NP-hard, involving an  exponential sum of terms
2. The exact computation of the derivative of the log-likelihood is  also NP-hard, since it contains the derivative of the partition function

Therefore, in practice we have to approximate both the computation of
the probabilities and several components of the learning process
itself. These drawbacks have prevented RBMs to show their real
potential as truly probabilistic models.

Currently, we are working on trying to improve several of the unsolved
issues related to RBMs:

1. - Mechanisms to control the learning process: www.lsi.upc.edu/~eromero/Publications/Downloads/2018-tnnls-stopcritRBM.pdf
2. Better approximations of the derivative of the log-likelihood: http://www.lsi.upc.edu/~eromero/Publications/Downloads/2019-nn-weightedCD.pdf
3. Efficient approximation of the partition function (work in progress)

These works have opened new lines of research, some of which can be
the topic of a Master's Thesis. The scope and degree of depth of the
work can be adapted to the estimated times to complete the Thesis. For
further details, contact Enrique Romero (eromero@cs.upc.edu).

This is precisely why companies such as Google have invested a lot of resources on exploring applications of GNN, some of which are as mediatic as Alpha Fold (recently solving the Protein-Folding problem), or used by Goolge Maps to make predictions of the expected traffic of a road. In this regard, the goal of this thesis is to continue investigating about the potential of GNNs when applied to the field of Computer Networks. Our goal is to provide GNN-based solutions that make computer networks run more autonomously and in a more efficient way.

A thorough prediction of specific parameters expected during a sailing race is a valuable information for a sailor, as it completely conditions his/her tactics during the race. With the aim of developing a decision support system valid for the Olympic Classes Sailing Venues, the following components are being developed and integrated together into one single web-based platform: 1. Wind component 2. Waves component 3. Oceanic current component 4. Boat Performance component. The present master thesis proposal is related with the wind component. The goal is to develop a methodology/procedure to analyse the 'recorded wind dataset' and recognize significant wind patterns, in other words, characteristic features of the wind speed and direction related to the other weather parameters of the day (air pressure, air and water temperature, cloud coverage, etc..) and to the geographical position of the specific racing area. A similar analysis should be performed on the 'weather prediction model dataset', to find correlations between predicted and measured values. This step is fundamental for the validation of the weather model that will be used daily during the Olympic Games.

Traditional classification schemes for wind patterns are based on meteorological experience and human interpretation of synoptic weather charts. However, it is very useful to have approaches based on clustering that can automatically induce wind patterns based on collected/predicted data, as well as the characteristic features of these patterns and their evolution through the day. Moreover, because data measuring is being performed in different locations of the race areas, these automatic clustering methods will be key to identify different behaviours of the wind depending on the area for the same pattern. All these will allow:

    -          A detailed analysis to determine the representativeness of the wind fields encountered in the measuring period, their frequency of occurrence, timing, rate of evolution, and transition probabilities.

    -          Consequently, a more thorough prediction of the conditions expected before a sailing race, which is, as mentioned, a highly valuable information for the sailor.

    During the previous master projects the clustering module has been completely developed, and it has been applied to data coming from weather prediction models and to data collected on the sea (for Tokyo 2020 Olympic Games). This has been done applying traditional clustering techniques. Also, since these data are recorded daily and present a time-series behaviour, more advanced and novel sequential clustering techniques have been applied to these data in order to extract more complex sequential wind patterns. The current project aims then at:

-      Improving the clustering modules, including the automatic computation of the optimum number of clusters

-      Applying additional machine learning techniques in order to take into account available key parameters, such as geolocation

-      Develop a methodology for comparing/combining the results obtained with both data sets (weather models and actual data)

-      Develop a prediction module that suggests the forecasted pattern and behaviour for a racing day given the weather model data and eventually the actual data early in this day

-      Generalising the environment so that it can be used not only in Marseille (the Olympic venue for Paris2024), but in any particular location where wind patterns are interesting and data are available

    The format of data collected on the sea is the one provided by the Austrian Sailing federation who is the owner of the on-board wind measurement system, and the format of weather prediction models dataset is the one used by operational meteorological centres, and suggested by the World Meteorological Organization, for the storage and the exchange of gridded weather and climate models fields (grib2 or NetCDF).

    Although this is a modality A thesis, interaction/collaboration with TriM company is highly possible depending on the evolution of the work.

The goal of this thesis is to enable the application of Deep Reinforcement Learning techniques for optimization problems in very large and complex scenarios. Inspired by the last related work from OpenAI (https://openai.com/blog/evolution-strategies/), we will explore time-efficient alternatives to train a Graph Neural Network based DRL agent on large optimization scenarios in a reasonable amount of time.

A more detailed description of the project can be found in

https://www.cs.upc.edu/~eromero/Downloads/Retina-TFM-Project-01.pdf

The project is proposed in collaboration with Javier Zarranz Ventura
(Institut Clínic d'Oftalmologia, ICOF, Hospital Clínic de Barcelona, and
Institut d'Investigacions Biomèdiques August Pi I Sunyer, IDIBAPS),
which would provide a large annotated database to develop the project. For further information, please contact Alfredo Vellido (avellido@cs.
upc.edu) or Enrique Romero (eromero@cs.upc.edu).

Gradient descend based RBM learning is a computationally expensive procedure that requires the computation of a probability normalization constant (called the Partition function) that involves the computation of an exponentially large number of terms. In order to bypass that, different approximations exist, the most celebrated one being the Contrastive Divergence (CD) algorithm [1].

In this project we want to explore an alternative RBM model based on archetype (ARC) learning, described in [2]. The student will have to implement both the standard CD and the ARC algorithms in CUDA, and compare their performance in common benchmarking datasets.

[1] "Training restricted Boltzmann machines: An introduction",
Asja Fischer and Christian Igel, Pattern Recognition 47 (2014) 25-39

[2] "The emergence of a concept in shallow neural networks"
Elena Agliari, Francesco Alemanno, Adriano Barra, Giordano De Marzo, Neural Networks 148 (2022) 232-253

 The goal of this thesis is to enable the application of Deep Reinforcement Learning techniques for optimization problems in very large and complex scenarios, in this case Data Center Networks. Inspired by the last related work from OpenAI (https://openai.com/blog/evolution-strategies/), we will explore time-efficient alternatives to train a Graph Neural Network based DRL agent on large optimization scenarios in a reasonable amount of time.

TinyML is a growing community that does machine learning on microcontrollers https://www.tinyml.org. 

Microcontrollers are sometimes the only choice when the power supply is limited, e.g. in for battery- or solar-powered applications. Application examples of TinyML are "wildlife" observation, such as: https://mybirdbuddy.com/ https://opencollar.io/, but tiny machine learning is also in domestic, healthcare and industrial applications.

While Arduino Uno boards are well known for all kinds of hobbyist microcontroller projects, for machine learning the more powerful 32 bit microcontrollers and development boards are used, such as the Arduino Portenta H7 or Arduino Nano 33 BLE Sense. These are able to run machine learning applications. To get the first information on this topic have a look at TensorFlow Lite for Microcontrollers: https://www.tensorflow.org/lite/microcontrollers.

In this project we would like to explore on-device machine learning model training specifically doing it by federated learning. We would like to use the Arduino Portenta H7, the most powerful microcontroller board from the Arduino family. We have Arduino Portenta H7 that can be used for the project. 

We have done previous work and code available where we showed that it is feasible.

https://www.mdpi.com/2079-9292/11/4/573

In the project it would be interesting to explore the capacities of the Portant H7, like using the two cores of the board, use at least one the sensors (camera, accelaration, microphone,...), and use the network connectivity (Wifi, BLE, LoRa)...

One concrete case would be to use for the communication part of federated learning the Wifi connectivity of the Arduino Portenta H7 to build a "cluster" (inspired by clusters of Raspberry Pi) where all the nodes are interconnected and can participate in federated learning. 

The project can start with already working code (in C/C++, Python) that we have from previous work on TinyML with on-device training and federated learning. https://upcommons.upc.edu/bitstream/handle/2117/353756/160036.pdf

The main objective of this project is to apply the most recent advances in Deep Learning, including Deep Reinforcement Learning and Graph Neural Networks, to uncover the particular methods used to track Internet users and collect PI.

This project will be useful for both Internet users and the research community, and will produce open source tools, real data sets, and publications revealing most privacy attempting practices.

Some preliminary results of our work in this area were recently published in Proceedings of the IEEE (IF: 9.237) and featured in a Wall Street Journal article.

More info at:

http://personals.ac.upc.edu/pbarlet/papers/web-tracking.survey2015.pdf

http://blogs.wsj.com/digits/2015/08/04/7-ways-youre-being-tracked-online-and-how-to-stop-it/