Research

Research Focus Areas

The work done by ICSE members in computational mathematics focuses on developing mathematical concepts, the derivation of constructive models, the creation of scalable computational methods, large-scale numerical simulations on supercomputers, and applications in various scientific and engineering fields.

Our team uses parallel computing, deep learning, and software/hardware co-design techniques to implement and optimize systems. The representative work is his important contribution in mobile wireless systems of the second and third generation.

Our researchers focus on quantifying and reducing statistical errors and risks, understanding of the uncertainty propagation through complex systems, and the derivation of effective computation machinery.

Researchers at ICSE are interested in settings that involve a high-dimensional dataset. For example, their work has produced efficient algorithms for various statistical estimation tasks, provided new algorithmic frameworks for solving interactive machine learning problems, and has led to the formation of scalable tools for machine learning applications.

Members of ICSE construct models based on inference, verification, and validation paradigm in many disciplines. For example, our team adapt current models so they can easily be solved in a depth averaged context, and the second is to implement robust and efficient solvers for the simulations. Besides, we developed a novel conceptual framework, called statistical teleodynamics, that synthesizes key concepts and techniques from artificial intelligence, game theory, statistical mechanics, systems engineering, economics, biology, and philosophy to understand and model various phenomena.

The ICSE members use multiscale theoretical, experimental, and numerical approaches to investigate various research frontiers in materials and mechanics addressing challenges in energy and environment, nanomechanics, and mechanobiology. The theories and methods will be implemented by large scale computational models for virtual experiments and manufacturing of composite materials.

Our team in computational climate field unravel and quantify the fundamental reactivity of gases relevant to energy conversion and atmospheres of Earth and other planets, the impact of turbulence on the transport of mass, the continental hydrologic cycle using multiscale modeling and big data (machine learning, remote sensing, high-resolution turbulent simulations) in the context of rising CO2 concentrations, the chemistry and dynamics of the stratosphere, Arctic and Antarctic climate change, and past and future climate impacts related to the Montreal Protocol.

Team members apply biological mechanisms in diagnostic and therapeutic medical applications, the field of soft tissue with a specific focus on the female reproductive system and pregnancy, the emerging field of synthetic biology, and are also interested in what happens in our brains when we make a rapid decision.

Our members Aurel A. Lazar is interested in computing with neural circuits (in silico) and reverse engineering the fruit fly brain and leading the development of NeuroNLP and NeuroGFX, two key FFBO applications, that enable researchers to use plain English to probe biological data that are integrated into NeuroArch and provide users highly intuitive tools to execute neural circuit models with Neurokernel.

Our team members focus on computer graphics algorithms that address various geometric and physical problems, such as predicting the motion and deformation of materials, processing 3D geometric data, and interactive tools for engineering design.

In computational health, we focus on the mathematical analysis and quantification of medical images, signal and image processing, computer-aided diagnosis, and biomedical informatics.

Our researchers in computational mechanics focus on the study of nanoscale structure and its role in the properties of diverse materials, the mechanical behavior and mechanics of materials at small scales and under extreme conditions, the computing materials behavior from the first-principles of quantum mechanics, the mechanics and physics of geological and porous materials, the Fracture and Damage Mechanics considering the Multiphysics and Multiscale response of materials, and the computational quantum mechanical studies of materials at extreme conditions, especially planetary materials.

ICSE members develop and apply materials simulation methods to study materials properties at high pressures and temperatures, especially minerals at the planetary interior condition. Other researches include water-ice physics and properties of strongly correlated oxides and their crystalline defects.

Our member Changxi Zheng develops computational technologies used for virtual reality, digital fabrication, and mechanical engineering. His work includes simulations of automatically generate realistic, complex motions, as well as the methods that compute virtual sounds synchronized with simulated motions.

Our team works on the theory and applications of optimization. We are interested in the interplay of mathematical programming and computer science to construct and solve mathematical models that support decision making. We also develop primarily analytic and numerical stochastic methodologies for response analysis, reliability assessment, and optimization of complex engineering systems and structures under the presence of uncertainties.

Members at ICSE work on random processes and computer simulation algorithms and develop tools so that data for one system can be used to make predictions about modified systems for which no data is available. Such tools are useful to regulators in predicting how well banks respond to financial shocks, to scientists in predicting future climate under various carbon-emissions rates, and to engineers in predicting how factory layouts will boost performance and solve decision problems in information-rich and highly dynamic environments.

Our researchers use regression architectures, machine learning techniques, and neural networks to develop unstructured and indeterministic approaches dealing with traditional mechanics of solid mechanics problems based on the powerful computational ability and a larger amount of material data.