Understanding the nature of representation in neural networks is a goal shared by neuroscience and machine learning. It is therefore exciting that both fields converge not only on shared questions but also on similar approaches. A pressing question in these areas is understanding how the structure of the representation used by neural networks affects both their generalization, and robustness to perturbations. In this work, we investigate the latter by juxtaposing experimental results regarding the covariance spectrum of neural representations in the mouse V1 (Stringer et al) with artificial neural networks. We use adversarial robustness to probe Stringer et al's theory regarding the causal role of a 1/n covariance spectrum. We empirically investigate the benefits such a neural code confers in neural networks, and illuminate its role in multi-layer architectures. Our results show that imposing the experimentally observed structure on artificial neural networks makes them more robust to adversarial attacks. Moreover, our findings complement the existing theory relating wide neural networks to kernel methods, by showing the role of intermediate representations.

Artificial intelligence can help accurately map and track penguin colonies in Antarctica by analysing tourist photos. "Right now, everyone has a camera in their pocket, and so the sheer volume of data being collected around the world is incredible," says Heather Lynch at Stony Brook University in New York. Haoyu Wu at Stony Brook University and his colleagues, including Lynch, used an AI tool developed by Meta to highlight Adélie penguins in photographs taken by tourists or scientists on the ground. With guidance from a human expert, the AI tool was able to automatically identify and outline entire colonies in photos. This semi-automated method is much faster than doing everything manually because the AI tool takes just 5 to 10 seconds per image, compared with a person taking 1 to 2 minutes, says Wu. The team also created a 3D digital model of the Antarctic landscape using satellite imagery and terrain elevation data.

Development of a Novel Impedance-Controlled Quasi-Direct-Drive Robot Hand by Jay Best Master of Science in Mechanical Engineering Stony Brook University 2023 Most robotic hands and grippers rely on actuators with large gearboxes and force sensors for controlling gripping force. However, this might not be ideal for tasks which require the robot to interact with an unstructured and/or unknown environment. We propose a novel quasidirect-drive two-fingered robotic hand with variable impedance control in the joint space and Cartesian space. The hand has a total of four degrees of freedom, a backdrivable gear train, and four brushless direct current (BLDC) motors. Field-Oriented Control (FOC) with current sensing is used to control motor torques. Variable impedance control allows the hand to perform dexterous manipulation tasks while being safe during human-robot interaction. The quasidirect-drive actuators enable the fingers to handle contact with the environment without the need for complicated tactile or force sensors. A majority 3D printed assembly makes this a lowcost research platform built with affordable off-the-shelf components. The hand demonstrates grasping with force-closure and form-closure, stable grasps in response to disturbances, tasks exploiting contact with the environment, simple in-hand manipulation, and a light touch for handling fragile objects.

Machine Learning for Probabilistic Prediction Quantitative Finance Webinar, Stony Brook University (11/11/2022) Valery Manokhin, PhD, MBA, CFQ Speaker Bio • PhD in Machine Learning (2022) from Royal Holloway, University of London • During PhD conducted research and published papers in probabilistic and conformal prediction. PhD supervised by Prof. Vladimir Vovk, the creator of Conformal Prediction (Prof. Vladimir Vovk is the last PhD student of Andrey Kolmogorov) • Dr. Valery Manokhin holds a number of advanced MSc degrees including from the Moscow Institute of Physics and Technology (Physics/Math), UCL (Computational Statistics and Machine Learning), University of Sussex (Quant Finance) and an MBA from the University of Warwick • Published in the leading machine learning journals, including'Neurocomputing', 'Journal of Machine Learning Research' and'Machine Learning Journal', also in the industry journals including'Frontiers in Energy Research' • Created'Awesome Conformal Prediction' - the most comprehensive professionally curated resource on Conformal Prediction (over 900 stars on GitHub). 'Awesome Conformal Prediction' has been featured at the leading conferences such as ICML and in Kevin Murphy's bestselling book'Probabilistic Machine Learning: An Introduction' Outline of this webinar Introduction to Probabilistic Prediction Probability Calibration Introduction to Conformal Prediction Conformal Prediction for Classification Conformal Prediction for Regression Conclusion 3 Why Probabilistic Prediction? Machine Learning is primarily concerned with producing functions mapping objects onto predicted labels Classical statistical techniques - for small scale, low-dimensional data High-dimensional data does not necessarily follow well-known distributions and hence required new approaches (e.g.

For the last two years, Facebook AI Research (FAIR) has worked with 13 universities around the world to assemble the largest ever data set of first-person video--specifically to train deep-learning image-recognition models. AIs trained on the data set will be better at controlling robots that interact with people, or interpreting images from smart glasses. "Machines will be able to help us in our daily lives only if they really understand the world through our eyes," says Kristen Grauman at FAIR, who leads the project. Such tech could support people who need assistance around the home, or guide people in tasks they are learning to complete. "The video in this data set is much closer to how humans observe the world," says Michael Ryoo, a computer vision researcher at Google Brain and Stony Brook University in New York, who is not involved in Ego4D.

The coronavirus disease (COVID-19) has rapidly spread throughout the world and while pregnant women present the same adverse outcome rates, they are underrepresented in clinical research. We collected clinical data of 155 test-positive COVID-19 pregnant women at Stony Brook University Hospital. Many of these collected data are of multivariate categorical type, where the number of possible outcomes grows exponentially as the dimension of data increases. We modeled the data within the unsupervised Bayesian framework and mapped them into a lower-dimensional space using latent Gaussian processes. The latent features in the lower dimensional space were further used for predicting if a pregnant woman would be admitted to a hospital due to COVID-19 or would remain with mild symptoms. We compared the prediction accuracy with the dummy/one-hot encoding of categorical data and found that the latent Gaussian process had better accuracy.

Stony Brook University professor Peng Zhang, a SUNY Empire Innovation professor in the Department of Electrical and Computer Engineering, is leading a statewide team of collaborators in developing "AI-Grid," an artificial intelligence-enabled, autonomous grid designed to keep power infrastructure resilient from cyberattacks, faults and disastrous accidents. The work is part of the National Science Foundation's (NSF) Convergence Accelerator Program, which supports and builds upon basic research and discovery that involves multidisciplinary work to accelerate solutions toward societal impact. In September 2020, the program launched the 2020 cohort, which included AI-Grid as a phase 1 awardee and grant funding of a $1 million to further AI-Grid research from an idea to a low-fidelity prototype. The Convergence Accelerator recently selected teams for phase 2, to focus on expanding the solution prototype and to build a sustainability plan beyond the NSF funding. Under phase 2, a new $5

Chest x-rays used in the COVID-Net study show differing infection extent and opacity in the lungs of COVID-19 patients. Artificial intelligence (AI) technology developed by researchers at the University of Waterloo is capable of assessing the severity of COVID-19 cases with a promising degree of accuracy. A study, which is part of the COVID-Net open-source initiative launched more than a year ago, involved researchers from Waterloo and spin-off start-up company DarwinAI, as well as radiologists at the Stony Brook School of Medicine and the Montefiore Medical Center in New York. Deep-learning AI was trained to analyze the extent and opacity of infection in the lungs of COVID-19 patients based on chest x-rays. Its scores were then compared to assessments of the same x-rays by expert radiologists.

As a global cornerstone of the artificial intelligence community, NeurIPS … "NeurIPS is the premier conference in machine learning, with prominence …

Research published recently by CSE investigators can make training machine learning (ML) models fairer and faster. With a tool called AlloX, Prof. Mosharaf Chowdhury and a team from Stony Brook University developed a new way to fairly schedule high volumes of ML jobs in data centers that make use of multiple different types of computing hardware, like CPUs, GPUs, and specialized accelerators. As these so-called heterogeneous clusters grow to be the norm, fair scheduling systems like AlloX will become essential to their efficient operation. This project is a new step for Chowdhury's lab, which has recently published a number of tools aimed at speeding up the process of training and testing ML models. Their past projects Tiresias and Salus sped up GPU resource sharing at multiple scales: both within a single GPU (Salus) and across many GPUs in a cluster (Tiresias).