74 young scientists from the Wrocław University of Science and Technology Doctoral School, including 8 from our Faculty, received funding for the implementation of their projects as part of the first edition of the competition prepared by our university. Mini-grants can be used, among others, for the purchase of equipment, research trips or training.
The program is the result of the efforts and work of the WUST Doctoral Students Council, which has been for a while reporting the need to introduce such a solution. The program will support the scientific activity of doctoral students and allow their more complex self-development in selected research fields.
The grant could be applied for by 2nd, 3rd and 4th year doctoral students (excluding implementation doctoral students) who were not project managers or managers of a research task in a research project during the last calendar year prior to the moment of submitting the application. Additionally, their education, as specified in the Individual Research Plan, should not end during the implementation of the grant.
The funding may be used, among others, for the purchase of equipment, devices, materials or reagents necessary to conduct research, covering the costs of a short-term research trip, participation in scientific conferences or in workshops or training.
A total of PLN 1,020,000 was allocated to support doctoral students in the competition.
Below are the 12 students whose applications received the highest rating.
In the discipline of information and communication technology, mini-grants were awarded to doctoral students from our Faculty:
Betelhem Wubineh is writing a thesis on: “Segmentation and Classification of Pap Smear Images to Detect Cervical Cancer Using Deep Learning Techniques.”
“The aim of my research is to perform semantic segmentation on images of cytological smears in order to detect cervical cancer,” says the doctoral student. “Specifically, the idea is to precisely separate cervical cells from the background using deep learning techniques, and thus to automate the segmentation of individual cells. This approach improves the identification and analysis of abnormal cellular structures associated with cervical cancer, providing valuable support for cytotechnologists in diagnosis and decision-making.”
Jose Fabio Ribeiro Bezerra is writing his doctoral thesis in the field of computer science and telecommunications.
“My research focuses on detecting false information on Twitter. For this purpose, I use both the content of the tweet and any additional information such as images and videos. The aim of the work is to develop a model that detects false information better than other models which use only text-based information,” explains the PhD student from the Department of Applied Computer Science.
He will use his grant to improve the efficiency of his research. He says he needs large amounts of data that a single laptop cannot process in a reasonable time. He will use the grant to build a separate computer cluster (a group of computers working in parallel) based on used hardware. “This equipment has been previously used by other companies, but is still in good condition. I will be able to process large amounts of data,” explains Jose Fabio Ribeiro Bezerra. “Reusing equipment that other people no longer find useful for their business also helps reduce the carbon footprint produced by such clusters.”
In his Ph.D. program, Jędrzej Kozal is working on life-long learning in machine learning. Recent years have seen significant progress in the field of machine learning through the use of deep models, allowing significant progress in many areas.
“Nevertheless, deep models have an undesirable property. When they learn a new task, they quickly forget the previously accumulated knowledge,” says Jędrzej Kozal. “In terms of teaching people, this would mean that someone who knows how to play tennis, would forget it after learning how to ride a bike. This phenomenon is called "catastrophic forgetting." The aim of long-term learning is to eliminate this phenomenon. Long-term learning can enable previously trained neural networks to be more easily adapted to new data and learn new tasks with limited forgetting of the previously acquired knowledge. This could allow existing networks to adapt to new data and changing environments.
The doctoral student from the Department of Computer Systems and Networks will spend the mini-grant on a trip to the USA. “There is a team of Professor Bartosz Krawczyk, who deals with continual learning. I managed to cooperate with this team earlier, which resulted in joint research on the use of interpolation of neural network weights in long-term learning,” explains Jędrzej Kozal. “A paper with the results of this research was accepted for the CLVision workshop at the CVPR conference. Now we want to continue cooperation, develop previous research and work on new ideas.”
Dominikia Kunc is a Ph.D. student at the Department of Artificial Intelligence and a member of the Emognition research team. She is preparing a dissertation under the supervision of Prof. Przemysław Kazienko and Prof. Stanisław Saganowski.
“In my Ph.D., I deal with learning the representation of physiological signals collected in everyday life using wearable devices, such as ECG wristbands or popular smartwatches,” says Dominika Kunc. “Together with the Emognition team, we have conducted a study of affect in everyday life, which allowed us to collect over 35,000 hours of recordings of physiological signals from 167 people along with over 18,000 affective markings. Representation learning transforms raw data into a more concise and informative form by using large amounts of unlabelled data in model training. Representations have the ability to capture hidden patterns and relationships in data, allowing their better understanding and interpretation. Unfortunately, the training of such models is time-consuming, due to among others the size of the data used.
As part of the grant, the doctoral student will purchase a computational server that will enable ethical data processing and conduct large-scale experiments in the local environment. She plans to perform two experiments. The first is an exploratory analysis of clusters detected in physiological data collected in everyday life to discover patterns and measure their variability within and between individuals, using the full physiological information contained in the representation. The second is to create a psychophysiological map of affect using data visualization and dimensionality reduction techniques on learned representations.
“Understanding and recognizing affective states can be a comprehensive tool supporting people with emotional disorders and improving their self-awareness and mental health,” adds Dominika Kunc.
Rajiansyah is working on a dissertation titled: The Divide and Conquer Strategy for Group Decision Making.
“The grant will help me to pay for presenting my topic at the conference, publish to some journal and doing some workshop,” says the Ph.D. student.
Katarzyna Białas is preparing her doctoral thesis on a multi-factor user authentication system based on the analysis of brain wave (EEG) signals using a brain-computer interface (BCI). Brain-computer interfaces are systems of direct communication between humans and machines, the main role of which is to translate the brain signal into the control signal. One of the areas in which brain-computer systems can be used is cybersecurity, in particular additional security of the IT system using the EEG brain signal analysis, i.e. electroencephalography.
“I will use the grant to participate in the international conference IEEE SMC 2024 (IEEE International Conference on Systems, Man, and Cybernetics), which takes place in October this year in Borneo, Malaysia,” says Katarzyna Białas. “Participation in this conference will provide me with direct access to the latest achievements in the use of machine learning algorithms and computational intelligence in the analysis of data generated by brain-computer interfaces. The conference includes special workshops ("Privacy Preserving Brain-Computer Interfaces") related to BMI, i.e. brain machine interaction. They will be an opportunity for me to learn about the latest achievements in the field of brain-computer interfaces and in particular about topics related to cybersecurity.
The doctoral student from the Department of Applied Computer Science adds that the “Neuron” Neuroinformatics Research Club from our Faculty is also involved in the topic of brain-computer interfaces and their multidimensional applications. Katarzyna Białas is one of the club's founders, and is currently its supervisor.
Aulia Arif Wardana: "My doctoral research aims to explore the integration of consensus algorithms with collaborative learning to enhance the intelligence, reliability, and trustworthiness of collaborative anomaly detection in Collaborative Intrusion Detection Systems (CIDS). By combining these technologies, the research seeks to mitigate the risks posed by malicious actors attempting to inject false or poisoned updates into the system. This research contributes to the field of information security and applications, particularly in the context of safeguarding modern networks against evolving threats".
"European Symposium on Research in Computer Security (ESORICS) is a prestigious symposium focused on computer security research in Europe, making it an ideal platform to enhance doctoral student understanding of cutting-edge research in the field. So, I'm going to spend this grant to help me attend ESORICS 2024. The symposium can provide valuable insights, networking opportunities, and exposure to the latest advancements in computer security. My dissertation topic is related to intrusion detection systems in computer security. By attending ESORICS, I can deepen my knowledge in these areas, stay updated on relevant research trends, and gain inspiration for further developing and finishing my doctoral thesis," explains the PhD student form our faclty.
Barbara Wędrychowicz, a Ph.D. student from the Department of Applied Computer Science, also received a mini-grant for her research. The topic of her doctoral dissertation is "Modelling the incrementation of knowledge in small diverse groups during communication."
“Transferring knowledge while working in groups is a problem in which several processes take place at the same time: communication, knowledge transfer, knowledge acquisition, as well as agreement and creation of knowledge,” says Barbara Wędrychowicz about her research. “Preparing a formal model requires taking into account both the possibility of simultaneous occurrence of the above-mentioned activities, and the diversity of agents and the impact of their features on selected processes. The use of multi-agent systems allows the developed formal model to be mapped and simulated."
The young researcher will use her grant to travel to a scientific conference. “The aim of this trip is to establish personal contacts with scientists dealing with topics related to the topic of my doctoral thesis,” says Barbara Wędrychowicz. During the conference, she plans to present both her current results and further research plans. She counts on feedback, substantive comments and tips that she will be able to use in further stages of her doctoral dissertation.
Congratulations to all young scientists!
Gallery
