## Programs & Events

##### Models and Machine Learning for Causal Inference and Decision Making in Health Research

Jan 14 - 18, 2019

Causal and mathematical models are widely used for decision making and policy evaluation at both the micro and macro levels. For example, causal models using large datasets are used to evaluate treatment efficacy in HIV; mathematical models are used to simulate the effects of prevention or policy measures to improve health outcomes or reduce the spread of infectious diseases. Entities such as the World Health Organization and UNAIDS rely on these models to set wide-ranging and high-impact policy related to treatment and prevention of infectious disease.

Causal models tend to rely on large-scale cohort data, while mathematical models in many ways represent evidence synthesis. Important methodologic issues in the development, application, and interpretation of these models include the role of untestable assumptions, transportability of findings to specific populations of interest, model calibration and validation, and uncertainty quantification. The datasets used to develop these... (more)

##### Organizing Committee

- Joseph Hogan

##### Scientific Machine Learning

Jan 28 - 30, 2019

The machine learning revolution is already having a significant impact across the social sciences and business, but it is also beginning to change computational science and engineering in fundamental and very varied ways.

We are experiencing the rise of new and simpler data-driven methods based on techniques from machine learning such as deep learning. This revolution allows for the development of radical new techniques to address problems known to be very challenging with traditional methods and suggests the potential dramatic enhancement of existing methods through data informed parameter selection, both in static and dynamic modes of operation. Techniques are emerging that allows us to produce realistic solutions from non-sterilized computational problems in diverse physical sciences.

However, the urgent and unmet need to formally analyze, design, develop and deploy these emerging methods and develop algorithms must be addressed. Many central problems, e.g., enforcement of... (more)

##### Organizing Committee

- Jan Hesthaven
- George Karniadakis

##### Algebraic Vision Research Cluster

Jan 28 - Feb 15, 2019

The powerful communication between mathematics and other scientific communities can be challenging. Notably, the interface between nonlinear algebra and computer vision has seen a boost of activity, giving rise to the Algebraic Vision community. At ICERM, this interface was jump-started at the Vision Working Group of the 2018 Nonlinear Algebra Program, when leading researchers came together to formulate new problems and solutions to push the field forward. We would like this activity to continue into the 2019 Computer Vision Program at ICERM, in the form of a research cluster, further bridging the fields to new levels.

##### Organizing Committee

- Ricardo Fabbri
- KathlĂ©n Kohn

##### Abelian varieties over finite fields

Jan 31 - Feb 3, 2019

This will be a hands-on workshop focused on a specific computational problem: enumerating all isomorphism classes of abelian varieties of of dimension g over a finite field of cardinality q, for a suitable range of integers g and q. Isogeny classes of abelian varieties over finite fields have been previously classified by Weil polynomials and can be found in the L-functions and Modular Form Database. The goal of this workshop is to refine this to the level of isomorphism classes, and, whenever possible, to construct explicit representatives for each isomorphism class. By exploiting recent theoretical and computational advances and assembling an appropriate team of experts, we hope to make rapid and substantial progress during this short, focused workshop and to have results available in advance of the conference on the Arithmetic of Low-dimensional Abelian Varieties that will take place at ICERM in June.

**This is a closed workshop that will not be accepting applications.**

##### Organizing Committee

- Andrew Sutherland
- John Voight

##### Computer Vision

Feb 4 - May 10, 2019

Computer vision is an inter-disciplinary topic crossing boundaries between computer science, statistics, mathematics, engineering and cognitive science.

Research in computer vision involves the development and evaluation of computational methods for image analysis. This includes the design of new theoretical models and algorithms, and practical implementation of these algorithms using a variety of computer architectures and programming languages. The methods under consideration are often motivated by generative mathematical models of the world and the imaging process. Recent approaches also rely heavily on machine learning techniques and discriminative models such as deep neural networks.

Problems that will be considered in the program include image restoration, image segmentation, object recognition and 3D reconstruction. Current approaches to address these problems draw on a variety of mathematical and computational topics such as stochastic models, statistical methods,... (more)

##### Organizing Committee

- Yali Amit
- Ronen Basri
- Tamara Berg
- Alex Berg
- Pedro Felzenszwalb
- Benar Fux Svaiter
- Stuart Geman
- Basilis Gidas
- David Jacobs
- Olga Veksler

##### Theory and Practice in Machine Learning and Computer Vision

Feb 18 - 22, 2019

Recent advances in machine learning have had a profound impact on computer vision. Simultaneously, success in computer vision applications has rapidly increased our understanding of some machine learning techniques, especially their applicability. This workshop will bring together researchers who are building a stronger theoretical understanding of the foundations of machine learning with computer vision researchers who are advancing our understanding of machine learning in practice.

Much of the recent growth in the use of machine learning in computer vision has been spurred by advances in deep neural networks. At the same time, new advances in other areas of machine learning, including reinforcement learning, generative models, and optimization methods, hold great promise for future impact. These raise important fundamental questions, such as understanding what influences the ability of learning algorithms to generalize, understanding what causes optimization in learning to converge... (more)

##### Organizing Committee

- Ronen Basri
- Alex Berg
- David Jacobs

##### An ICERM Public Lecture: Discovering Black Holes and Gravitational Waves: Algorithms and Simulation

Feb 20, 2019

The equations of general relativity, Einstein's field equations, are among the most complicated partial differential equations in mathematical physics. These equations predict the existence of gravitational waves, which are propagating disturbances in spacetime itself. In 2016, the first direct observation of these waves from colliding black holes was reported on, a historic discovery that led to last years Nobel Prize. This discovery would not have been possible without intense interaction between physicists, mathematicians, and high-performance computing tools. Indeed, numerically solving Einstein's equations for the expected wave signal and the processing of gravitational-wave datasets was enabled by advances in algorithms, numerical methods, and access to large computing resources. In this talk, I will focus on the critical role all three played in making this historic discovery as well as summarizing current directions in computational relativity and gravitational-wave data... (more)

##### Image Description for Consumer and Overhead Imagery

Feb 25 - 26, 2019

Building systems that can understand visual concepts and describe them coherently in natural language is fundamental to artificial intelligence. Advances in machine learning have had profound impact on computer vision and natural language processing. There has been interesting progress in recent years at the intersection of these two fields, producing systems that describe (eg., caption) images and videos captured by personal cameras in ordinary scenes and street views. Much work remains in this and a host of related problems, including that of building natural language descriptions of commercial overhead imagery and videos, where automation is greatly needed: "If we were to attempt to manually exploit the commercial satellite imagery we expect to have over the next 20 years, we would need eight million imagery analysts" [Robert Cardillo, NGA Director, GEOINT Symposium 2017]. This workshop brings together researchers in machine learning, computer vision, natural language processing... (more)

##### Organizing Committee

- Trevor Darrell
- David Jacobs
- Triet Le
- Guillermo Sapiro
- Eric Xing

##### Modularity and 3-manifolds

Mar 8 - 10, 2019

A long-standing problem in quantum topology is to find a function, more precisely a q-series with integer coefficients, such that its limiting values at primitive roots of unity yield invariants of Witten and Reshetikhin-Turaev. In other words, such a function would be to 3-manifolds what the Jones polynomial is to knots. Somewhat surprisingly, recent physics developments suggest that, in order to solve this problem, one must associate to a 3-manifold not a single function (q-series), but rather a collection of functions. Very recently, based on both physical intuition and explicit computations, it was suggested that these 3-manifold invariants display (modified) modularity properties of various types and are related to number theoretic objects including mock and false theta functions and quantum modular forms. This workshop will bring experts from the fields of topology, physics and number theory together, with the goal of combining knowledge and computational skills and furthering... (more)

##### Organizing Committee

- Miranda Cheng
- Sergei Gukov

##### Computational Imaging

Mar 18 - 22, 2019

Computational imaging involves the use of mathematical models and computational methods as part of imaging systems. Algorithms for image reconstruction have important applications, including in medical image analysis and imaging for the physical sciences. Classical approaches often involve solving large inverse problems using a variety of regularization methods and numerical algorithms.

Current research includes the development of new cameras and imaging methods, where the hardware system and the computational techniques used for image reconstruction are co-designed. New developments have been influenced by the introduction of novel techniques for compressed sensing and sparse reconstruction. The use of machine learning methods for designing a new generation of imaging systems has also been increasingly important.

Specific topics that will be discussed include: image reconstruction, computational photography, compressed sensing, machine learning methods, numerical optimization,... (more)

##### Organizing Committee

- Pedro Felzenszwalb
- Basilis Gidas

##### An ICERM Public Lecture - Bias in bios: fairness in a high-stakes machine-learning setting

Mar 21, 2019

Machine learning algorithms form biases, like humans, based on the data they observe. However, unlike humans, the algorithms can readily admit their biases when probed appropriately. Using publicly available lists of names, we enumerate biases in an unsupervised fashion from word embeddings trained on public data. Gender, racial, and religious biases emerge, among others. We then analyze the effects of these biases on a problem motivated by recommending jobs to candidates. To collect data for this task, we extract hundreds of thousands of third-person bios from the web. The straightforward application of machine learning is found to amplify some biases. However, unlike humans, it is easy to put in place algorithmic corrections to mitigate this bias amplification.

Joint work with: Maria De Arteaga (CMU); Alexey Romanov (UMass Lowell); Nat Swinger (Lexington HS); Tom Heffernan (Shrewsbury HS); Christian Borgs, Jennifer Chayes, and Hanna Wallach (MSR); Alex Chouldechova (CMU; Mark... (more)

##### Optimization Methods in Computer Vision and Image Processing

Apr 29 - May 3, 2019

Optimization appears in many computer vision and image processing problems such as image restoration (denoising, inpainting, compressed sensing), multi-view reconstruction, shape from X, object detection, image segmentation, optical flow, matching, and network training. While there are formulations allowing for global optimal optimization, e.g. using convex objectives or exact combinatorial algorithms, many problems in computer vision and image processing require efficient approximation methods.

Optimization methods that are widely used range from graph-based techniques and convex relaxations to greedy approaches (e.g. gradient descent). Each method has different efficiency and optimality guarantees. The goal of this workshop is a broad discussion of mathematical models (objectives and constraints) and robust efficient optimization methods (exact or approximate, discrete or continuous) addressing existing issues and advancing the state of the art.

##### Organizing Committee

- Yuri Boykov
- Pedro Felzenszwalb
- Benar Fux Svaiter
- Olga Veksler