Interview with Dr. Tomasz Krzywicki—Mathematics Exceptional Reviewer
Name: Dr. Tomasz Krzywicki
Affiliation: Chair of Mathematical Methods of Computer Science, Faculty of Mathematics and Computer Science, University of Warmia and Mazury, Poland
Research Interests: artificial intelligence in medicine; computer vision; image processing; machine learning; parallel processing; reinforcement learning
The following is a short interview with Dr. Tomasz Krzywicki:
1. Could you give a brief introduction of yourself to the readers? Could you introduce your current research direction and provide an update on your progress?
Both my research profile and education are related to computer science. In 2020, I graduated from the University of Warmia and Mazury in Olsztyn, Poland, with a degree in computer science. Since then, I have been working there continuously, first as a research assistant and later as an assistant professor. I have been conducting scientific research since my student days, focusing on the development and application of image processing and recognition methods. From the very beginning of my academic career, I have been collaborating with an interdisciplinary group consisting of computer scientists and ophthalmologists. We are developing novel methods to support the clinical diagnosis of ophthalmic diseases, while also examining existing methods and intelligent systems. The results of this research include the development of new methods for the ensemble localization of key anatomical structures in color retinal fundus photographs. My research also focuses on purely computer science aspects, where I devote a significant portion of my attention to modelling uncertain data. This is a crucial aspect in the current era of big data, where the costs of labelling data are often significantly higher than the costs of obtaining it. In addition to my academic work, I have been actively involved in the commercial IT industry for 10 years as a senior software developer and machine learning research engineer. I have collaborated on several projects with leading global pharmaceutical corporations. Currently, I am involved in the aviation sector in this field.
2. Can you please share with us your sentiments upon winning the award?
Peer review is a process in which researchers from the global community share their time, knowledge, and experience to express their opinions on new scientific research results. It is an activity that contributes to the quality, reliability, and credibility of global science. Peer review itself is not, and should not be, self-serving; therefore, receiving the news of my nomination for this prestigious award came as a massive surprise to me. It is very nice to be recognized in this way, especially since I never expected any tangible benefits from it. The scientific literature is similar in nature to the concept of open source in software, and ensuring its quality should be a shared interest of the entire academic community.
3. Could you share some insights into your approach to reviewing manuscripts? How do you balance thoroughness with efficiency?
Due to the short deadline for submitting reviews, the efficiency of the process itself must be high. First of all, I do not wait until the last minute and try to familiarize myself with the general outline of the article immediately after accepting the invitation to review. The topics of some papers are not as closely related to my areas of research interest, so I also reserve a day or two to explore the current state of knowledge on the topic itself. In many cases, however, due to significant thematic discrepancies, I have been forced to decline such invitations. In most cases, thoroughly reading the article and noting down all my comments and observations is a process that takes several hours to complete. However, in the case of some papers, especially those on complex topics, this time could be extended to several days. The most important conclusion from my experience is to give yourself time to calmly and thoroughly familiarize yourself with the research description. Rushing not only makes it difficult to focus on a topic that often involves deep abstractions, but can also cause you to overlook even obvious but significant methodological or lexical errors. The aforementioned short deadline for submitting a review may necessitate a compromise between thoroughness and efficiency; however, the scales should still tip in favor of thoroughness due to its contribution to ensuring quality.
4. What are the key factors and aspects that you consider most when reviewing a manuscript?
Each field of science has its own framework, both in terms of research topics and the way they are communicated to the global community. One of the pioneers in describing scientific experiments was Galileo, who divided them into three parts: formulation of the problem/research hypothesis, description of the experiment and observations, and conclusions with generalizations. As I come from technical sciences, this scheme is very familiar to me. Of course, it has evolved significantly over the years, along with the development of scientific disciplines. However, certain constants should be included in such an article, namely a specific research problem or hypothesis, methodology, description of results, discussion and conclusions resulting from the entire experiment. In experimental computer science, it is imperative that readers can easily reproduce the experiment, so in addition to a detailed description of the methodology, tools used, quality indicators, precise results, and conclusions from the authors, it is also essential to include implementations of the tools used to obtain these results, or the data used in the development and testing of the presented methods. Unfortunately, I have often encountered serious methodological errors in articles, where the authors' assumptions were entirely at odds with generally accepted standards. I have also come across papers where the authors refused to test the proposed methods on publicly available data. Such practices significantly undermine the credibility and quality of the research described, so these are key aspects to which I pay the most attention. When analyzing scientific papers, I also try to catch any incorrectly used phrases, missing and unclear passages, descriptions of mathematical equations and illustrations, and all other incomplete phrases that may mislead the reader. Scientific articles are a special type of publication that should be prepared with particular care.
5. Based on your experience, which research topics do you think will be of particular interest to the research community in the coming years?
Computer science has repeatedly been recognized as the fastest-growing scientific discipline in recent years. This trend is particularly evident in the field of heuristics, especially in the context of neural networks and deep learning. These methods have revolutionized not only the popularity of the research topic itself, but also its applications in other scientific disciplines and in everyday life. The trend of peak popularity of topics related to artificial intelligence achieved through neural networks is unlikely to change in the near future. While neural networks themselves are no longer experiencing the same dynamic growth as they did a few years ago, their ability to be adapted to solve almost any problem will keep them at the forefront for a long time to come. It is impossible not to mention that the fascination with artificial intelligence dictates the popularity of neural networks and deep learning. In the slightly more distant future, the podium of the most popular areas of research will be occupied by methods that are the successors of the aforementioned neural networks.