Search by tag «Algorithm» 4 results

• ITMO’s Andrey Filchenkov: How To Train Neural Networks To Identify Debtors Based on Their Expenses

  Researchers at ITMO’s Information Technologies and Programming Faculty created a model that can tell responsible loaners from those who don’t pay back their bank loans by the spending pattern. The algorithm can also serve as the basis for a system that would notify credit card users about their risky behaviour. ITMO.NEWS asked Andrey Filchenkov, associate professor at ITMO, head of the Machine Learning Lab, and one of the authors of the project, how the algorithm works, if it is possible for it to make a mistake, how ethical it is to use it, and, most importantly, why people keep viewing neural networks apprehensively despite their obvious successes. 

  13.05.2021

• ITMO PhD Student Karina Pats on Working in France During the Pandemic

  Karina Pats, a PhD student at ITMO’s Faculty of Information Technologies and Programming, studies the functioning of nuclear receptors using molecular dynamics and docking. She works actively together with the laboratory of Ferdinand Molnár at Nazarbayev University in Kazakhstan where she did her internship last spring. Thanks to the Erasmus+ program, now she got to work on a bioinformatics algorithm project at the University of Lorraine, Nancy (France). ITMO.NEWS talked to Karina Pats to learn more about life in France, new projects and colleagues, as well as landing an internship during the pandemic.

  22.03.2021

• ITMO Scientists Develop an Algorithm That Makes the Inference of Population History Models for People and Animals Easier and More Accurate

  Bioinformatics scientists from ITMO University have developed a software tool, which quickly and effectively analyses genome data to infer probable models of demographic history of specific species, such as plants, animals or humans. Using complex calculations, the software can suggest with a very high degree of probability the course of history of a particular group of living organisms.  This includes identification of periods with mass extinction or mass population growth and information about contacts with other populations of the same species in the past. The scientists’ article about this methodology has been published in the journal Giga Science.

  04.03.2020

• Gennady Korotkevich Wins 2019 Topcoder Open Competition in Two Tracks

  The results of this year’s Topcoder Open Competition have recently been announced in Houston. The competition brings together programmers and developers from all over the world and consists of six tracks, including Marathon and Algorithm. These are the tracks that ITMO University graduate and PhD student Gennady Korotkevich took first place in, thus becoming the first participant who came first in two tracks at once.

  17.11.2019