researchers worldwide are doing applied research using artificial intelligence with the help of matlab and simulink in collaboration with mathworks. explore featured researchers, project topics, and related publications.
lead researcher
machine learning is a major part of dr. gleyzer’s research. he develops methods and algorithms and then applies them to particle physics challenges, such as particle identification, event reconstruction, and extraction of interesting physics in large hadron collider (lhc) data. his goal is to use data from the highest-energy and cosmic frontier experiments, such as the lhc and high-luminosity lhc and the rubin observatory, together with sophisticated machine learning algorithms to make quantitative statements about the possible existence of new physics, such as dark matter.
mathworks is supporting this research collaboration.
s. gleyzer, p. seyfert, and s. schramm, computational physics, may 19, 2019 (paper preprint)
m. andrews, j. alison, s. an, p. bryant, b. burkle, s. gleyzer, m. narain, m. paulini, b. poczos, e. usai, high energy physics - experiment, october 2020
lead researcher
dr. kovscek is interested in the chemistry and physics of unconventional geological formations (e.g., shale) because of their importance in the transition of energy processes to net-zero carbon emissions. aside from being a significant energy resource, unconventional formations serve as geological seals of conventional subsurface formations that may be used to sequester carbon dioxide, store intermittent renewable energy (e.g., green hydrogen), and isolate nuclear waste. unconventional formations are ubiquitous throughout the subsurface, but our engineering science knowledge is poor due to their nanoporous structure and extreme heterogeneity. his research group examines the fabric of porous media as well as the physics of flow at length scales that vary from the pore to the laboratory to the reservoir. the organizing themes are flow imaging to delineate the mechanisms of transport in porous media (water and gas) and the synthesis of models from experimental, theoretical, and field data.
building an image super resolution pipeline for electron microscopy image prediction of shale rock
mathworks is supporting this research collaboration.
k.m. guan, t.i. anderson, p. creux, a.r. kovsceka, computers & geosciences, volume 156, november 2021
t.i. anderson, b. vega, j. mckinzie, s.a. aryana, a.r. kovscek, nature: scientific reports, october 21, 2021
lead researcher
big data and ever-increasing computational power mean that in many areas of business and science, ai has become a powerful tool for inference and classification. however, for areas such as the physical sciences, especially related to climate and sustainability problems, data can be hard to produce and manage. using the concept of physics-informed machine learning, dr. raymond’s research has ventured into the design of novel biomedical devices, improving the detection of mild traumatic brain injury, and refocusing data and simulation uses on ocean health and biodiversity.
mathworks is supporting this research collaboration.
samuel j. raymond, david j. collins, richard o’rorke, mahnoush tayebi, ye ai & john williams, nature: scientific reports, may 26, 2020
samuel raymond, nicholas j. cecchi, hossein vahid alizadeh, ashlyn a. callan, eli rice, yuzhe liu, zhou zhou, michael zeineh, david b. camarillo, annals of biomedical engineering, february 2022 (doi: )
lead researcher
at the umass icenet [internal combustion engine network] consortium, we develop predictive machine learning algorithms to improve engine design cycles using computer models, allowing manufacturers to optimize their designs and create fuel-efficient engines. this collaborative venture utilizes the disruptive technology of ai, applying it to the engine research and development paradigm. some of our current research areas include improving turbulence model fidelity, developing better sub-models for spray/gas interactions, and using artificial neural networks for making combustion calculations faster.
icenet utilizes the disruptive technology of ai, applying it to the engine research and development paradigm.
mathworks is supporting this research collaboration.
g. portwood, p. mitra, m. ribeiro, d. schmidt, neurips 2019, second workshop on machine learning and the physical sciences, november 12, 2019
lead researcher
short-term heat demand forecasting in district heating (dh) systems is essential for a sufficient heat supply and optimal dh operation. in this study, a machine learning–based multistep short-term heat demand forecasting approach using the data of the largest slovenian dh system is considered. the proposed approach involved feature extraction and comparative analysis of representative machine learning–based forecasting models. nonlinear models performed better than linear models, and the best forecasting results were obtained by gaussian process regression. this model was selected to implement the online forecasting solution for the dh of ljubljana and has been generating forecasts with a mean absolute normalized error of 2.70% since november 2019.
p. potočnik, p. škerl, and e. govekar, energy and buildings, vol. 233, 2021
lead researcher
the speech disorders & technology lab (sdtl) is dedicated to developing assistive speech technologies, including silent speech interface and speech-driven brain-computer interface, as well as to conducting basic science research on neurogenic motor speech disorders and underling neurological mechanisms for speech communication. advanced computing techniques (e.g., machine learning) are heavily used in these projects.
ut austin researchers used matlab to derive whole phrases from meg signals as a first step toward developing a brain-computer interface that would enable als patients to communicate.
d. dash, p. ferrari & j. wang, frontiers in neuroscience, 14(290), 1–15, 2020
lead researcher
research at the gene center munich focuses on genome and systems biology, innate immunity and infection biology, and translational medicine. our approaches are often interdisciplinary and range from structural and molecular biology to computational and systemwide studies. our research teams want to understand how proteins and macromolecular complexes function at the structural and mechanistic levels. a particular focus is the high-resolution structural analysis of proteins and macromolecular complexes with cryo-electron microscopy, but also x-ray crystallography and other state-of-the-art technologies.
fully automated quality control and image categorization for determining ice thickness for cryo-electron microscopy grids
a powerful method for solving three-dimensional structures of biological macromolecules
mathworks is supporting this research collaboration.
lead researcher
prof. kumbasar’s research interests include artificial intelligence; computational intelligence (type 1 to type n fuzzy logic, shallow to deep neural networks, and local to global optimization); machine learning, including supervised, unsupervised, and reinforcement; conventional to intelligent control systems; and uav to agv autonomous systems.
development and deployment of deep learning–based type 2 fuzzy systems
k. bölat, t. kumbasar, 2021 ieee international conference on fuzzy systems (fuzz-ieee), july 11–14, 2021