Intelligent Technologies Research Group

Research theme

Research theme "Theoretical and applied aspects of machine learning and mathematical modelling (2019–2023)"
   Physical Sciences (Informatics) 41-T-12, Technological Sciences (Informatics Engineering) 42-E-4

Main goal

Develop and explore models based on machine learning and mathematical modelling for cyber security, anomalous data detection, real-time data classification, prediction, and explanation of results.

  Tasks in 2023:

• The suitability of deep learning algorithms for hyper-spectral data analysis.
• Image similarity evaluation study using deep Siamese neural networks.
• Development of fraud site detection algorithms.
• A review of gene mutation prediction algorithms.
• Overview of propaganda detection algorithms.
• An application of semantic analysis and machine learning algorithms for password guessing.
• Content recognition in digitally structured documents. Study of inequalities for zeta functions.
• Study of limit theorems for combinatorial numbers.
• Efficient algorithms for visualization of Mandelbrot shells associated with the Ryman zeta function.
• Securities price forecasting and investment portfolio formation using machine learning algorithms.
• Geometric topology, number theory and algorithms.

 Tasks in 2022:

• Applicability of machine learning algorithms to hyper-spectral data development for analysis.
• Examination of image similarity using Siamese neural networks.
• Modelling the behaviour of social systems in critical conditions.
• Imitative modelling of social cohesion and radicalization phenomena.
• Application of text semantic analysis and machine learning algorithms for guessing passwords.
• Context recognition in digitally structured documents.
• Prove limit theorems for numbers in the class of triangular arrays.
• Limit theorems for mixed-type open service networks.
• Open service networks operating under low load conditions.
• Modelling and simulation of social behaviour phenomena.

 Tasks in 2021:

• Application of machine learning for hyperspectral image analysis.
• GAN networks for phishing URL detection.
• Modeling multi-phase systems.
• Investigation of open computer networks when network load is small.
• Stable modeling in finance engineering.
• Practical aspects of Central limit theorem Borvein algorithm

 Tasks in 2019-2020:

• Application of machine learning in anomalous network data detection
• Adapting heavy tail models to identify abnormalities in security tasks of computer networks
• Modification of classification algorithms for CERN data certification
• Critical region inequalities for Riemann zeta function modules
• Create a recursive algorithm for the recognition of the parameters of the Wiener process monitored with noise
• Create a recursive algorithm for multi-dimensional Dirichlet distribution parameter recognition
• Testing and matching metrics and an agent-based model for measuring the impact of cultural processes on social capital to create a functional simulation prototype
• Get the primary results in the theory of networks with different performance nodes