Past events

Have all your questions about the Master of Science in Software Engineering (MSSE) program answered by attending this online information session.

RSVP now to reserve your spot.

Attendees will be sent a link prior to the event.
 

The computer science colloquium takes place on Mondays from 11:15 a.m. - 12:15 p.m.

This week's speaker, Rafael Ferreira da Silva (Oak Ridge National Laboratory), will be giving a talk titled "The Role of Scientific Workflows in Modern Computational Science: State-of-the-art and Challenges."

Abstract

Scientific workflows have been used almost universally across scientific domains, and have underpinned some of the most significant discoveries of the past several decades. Many of these workflows have high computational, storage, and/or communication demands, and thus must execute on a wide range of large-scale platforms, from large clouds to upcoming exascale high-performance computing (HPC) platforms. In this talk, I will provide a view of the state of the art and some of my previous research and technical contributions, and identify crucial research challenges in the workflows community. (https://arxiv.org/abs/2110.02168).

Biography

Rafael Ferreira da Silva is a Senior Research Scientist in the National Center for Computational Sciences (NCCS) at the Oak Ridge National Laboratory (ORNL). Prior to joining ORNL, Dr. Ferreira da Silva was a Research Assistant Professor in the Department of Computer Science at University of Southern California (USC), and a Research Lead at the USC Information Sciences Institute. Dr. Ferreira da Silva's research focuses on the efficient and resilient execution of large-scale scientific workflow applications on heterogeneous distributed systems, and the modeling and simulation of parallel and distributed computing systems. He has extensive experience leading/working on large-scale projects related to distributed computing platforms, cyberinfrastructure systems, and applications. He also brings substantial experience on community engagement and workshop organization. He was PI on 11 NSF-funded projects, and his scholarly work includes more than 80 peer-reviewed journal articles, conference proceedings articles, book chapters, editorials, and conference abstracts. He is currently the chair of the Workshop on Workflows in Support of Large-Scale Science (WORKS, held in conjunction with SuperComputing). He received his Ph.D. in Computer Science from INSA-Lyon, France, in 2013.

This week's speaker, Dongyeop Kang, will be giving a talk titled "Human and Data in the loop of NLP Pipeline."

Abstract

NLP systems trained on standard machine learning pipelines; annotation, learning, and evaluation, are limited to causing various problems; for instance, the dataset collected from crowd workers often contains annotation artifacts or repeating patterns; as the systems are deployed to real-world users, they are not well controlled, interpreted, or interacted with real users. To address these problems caused by the ML pipeline, I will discuss recent work from the Minnesota NLP group on human-centric and data-centric approaches. For the human-centric aspect, we collect human perception on linguistic styles and then make the model to mimic how humans perceive styles. Then we develop interactive NLP systems that help scholars better read and write academic papers. In the data-centric NLP, we model data informativeness based on various training dynamics and then use them to find new important data points for data augmentation and annotation. We believe more involvement of humans and consideration of data dynamics transforms the traditional ML-driven NLP pipeline to be more robust, interactive, and information-effective.

Biography

Dongyeop Kang is an assistant professor in the Computer Science Engineering department at the University of Minnesota, Twin Cities. He leads the Minnesota Natural Language Processing (NLP) group that aims to develop human-centered language technologies. His group's research lies at the intersection of computational linguistics, machine learning, and human-computer interaction.

He completed a postdoc at the University of California, Berkeley, and obtained a Ph.D. in the Language Technologies Institute of the School of Computer Science at Carnegie Mellon University. During his Ph.D. study, he interned at Facebook AI research, Allen Institute for AI (AI2), and Microsoft Research. He has been awarded the AI2 fellowship, CMU Presidential fellowship, and ILJU Ph.D. fellowship.

The UMN Machine Learning Seminar Series brings together faculty, students, and local industrial partners who are interested in the theoretical, computational, and applied aspects of machine learning, to pose problems, exchange ideas, and foster collaborations. The talks are every Thursday from 12 p.m. - 1 p.m. during the Fall 2021 semester.

This week's speaker, Steven L. Brunton (University of Washington), will be giving a talk titled "Machine Learning for Sparse Nonlinear Modeling and Control."

Abstract

Biography

Steven L. Brunton is a Professor of Mechanical Engineering at the University of Washington. He is also Adjunct Professor of Applied Mathematics and Computer science, and a Data Science Fellow at the eScience Institute. Steve received the B.S. in mathematics from Caltech in 2006 and the Ph.D. in mechanical and aerospace engineering from Princeton in 2012. His research combines machine learning with dynamical systems to model and control systems in fluid dynamics, biolocomotion, optics, energy systems, and manufacturing. He is a co-author of three textbooks, received the University of Washington College of Engineering junior faculty and teaching awards, the Army and Air Force Young Investigator Program (YIP) awards, and the Presidential Early Career Award for Scientists and Engineers (PECASE).

Give to the Max Day 2021 is Thursday, November 18.

This year, help CS&E support tomorrow’s leading programmers and problem solvers with a gift to Code the Gap.

Code the Gap is a University of Minnesota student group that partners with Twin Cities public middle and high schools to teach yearlong programming courses to girls and students from other underrepresented groups in computer science fields.

During its inaugural 2020-2021 school year, ten student volunteers from Code the Gap taught Python and other computing fundamentals twice a week over Zoom to students at the Community School of Excellence in St. Paul. This fall, the program has taken on 35 more volunteers, allowing the program to expand to a second partner school, Murray Middle School. Courses are now taught in-person, and the group is making plans to bring underrepresented students to the Twin Cities campus for coding classes, study sessions, hackathons, and visits from guest speakers who will share their stories about what led them to careers in STEM. Code the Gap also hopes to provide Chromebooks to students without access to a personal computer.

This Give to the Max Day, please consider supporting Code the Gap as they inspire a love of coding and problem solving in students who might otherwise feel shut out of a career in computer science or under-exposed to tech and STEM fields. Your gift to Code the Gap today will help cover the costs of the organization’s increased programming, including transportation fees to bring students to campus, the purchase and upkeep of hardware and classroom materials like Chromebooks and robotics kits, and costs associated with bringing speakers to campus and hosting other coding education events.

Donate to Code the Gap now

The last day to cancel full semester classes without college approval and receive a "W" is Monday, November 15.

View the full academic schedule on One Stop.
 

The computer science colloquium takes place on Mondays from 11:15 a.m. - 12:15 p.m.

This week's speaker, Jeremy Siek (Indiana University), will be giving a talk titled "Efficient Gradual Typing."

Abstract

Gradual typing combines static and dynamic typing in the same program. Siek et al. (2015) describe five criteria for gradually typed languages, including type soundness and the gradual guarantee. A significant number of languages have been developed in academia and industry that support some of these criteria (TypeScript, Typed Racket, Safe TypeScript, Transient Reticulated Python, Thorn, etc.) but relatively few support all the criteria (Nom, Gradualtalk, Guarded Reticulated Python). Of those that do, only Nom does so efficiently. The Nom experiment shows that one can achieve efficient gradual typing in languages with only nominal types, but many languages have structural types: function types, tuples, record and object types, generics, etc.

In this talk, I present a compiler, named Grift, that addresses the difficult challenge of efficient gradual typing for structural types. The input language includes a selection of difficult features: first-class functions, mutable arrays, and recursive types. I show that a close-to-the-metal implementation of run-time casts inspired by Henglein's coercions eliminates all of the catastrophic slowdowns without introducing significant average-case overhead. As a result, Grift exhibits lower overheads than those of Typed Racket.

Biography

Jeremy Siek is a Professor at Indiana University. Jeremy's interests include programming language design, type systems, mechanized theorem proving, and compilers.  Jeremy's Ph.D. thesis explored foundations for constrained templates, aka the "concepts" proposal for C++.  Prior to that, Jeremy developed the Boost Graph Library, a C++ generic library for graph algorithms and data structures. Jeremy post-doc'd at Rice University where he and Walid Taha developed the idea of gradual typing: a type system that integrates both dynamic and static typing in the same programming language.  Jeremy taught at the University of Colorado for many years and then moved to Indiana University. Jeremy is currently working with his student on the Grift compiler, which demonstrates that gradually typed programs can be compiled into efficient executables using the latest techniques in cast compression. Jeremy is also investigating the use of denotational semantics to prove the correctness of compilers proofs mechanized in Agda.  In 2009 Jeremy received the NSF CAREER award to fund his project: "Bridging the Gap Between Prototyping and Production". In 2010 and again in 2015, Jeremy was awarded a Distinguished Visiting Fellowship from the Scottish Informatics & Computer Science Alliance. From 2015 to the present Jeremy has been working with colleagues at Northeastern, Brown, and Maryland on the NSF-funded projects Gradual Typing Across the Spectrum and Performant Sound Gradual Typing.

Paul Stachour will be giving a talk titled "Proper System and Software Compartmentalization for Malware Prevention."

This session will also be available for live viewing on Zoom and will be recorded and posted on the UMSEC YouTube channel after the event.

Abstract

A brief history of compartmentalization in computing with discussion on why it is important, what techniques have worked well in the past, why those techniques are no longer used, and some possibilities for the future. Proper compartmentalization totally prevents malware and ransomware.

Biography

Paul Stachour is a Software Engineer who specializes in total life-cycle Software Quality Assurance. He has extensive expertise in finding large numbers of different kinds of important defects in software. In addition, he has expertise in creating processes to build, document, and track requirements, designs, code, tests, approvals, and software issues and is adept at creating correct, reliable, functional software in effective and efficient ways in many programming languages.

Paul has worked for computer manufacturers IBM, Honeywell, and Secure Computing. His industry experience includes Delphax printers, Boston Scientific Medical, Honeywell Aerospace, Net Perceptions Recommendations, and Detector Electronics fire detection systems. He has taught computing subjects at five different universities, including the University of Minnesota.

This week's speaker, Parikshit Maini, will be giving a talk titled "Mobile Robots in Agriculture."

Abstract

The presentation will start with an overview of some of the recent projects and research in agricultural robotics in the Robotic Sensor Networks Lab. Dr. Parikshit will then talk about our recent work on weed removal in organic dairy pastures using autonomous robots. The carbon footprint of using diesel-run farm vehicles for weed removal and other agricultural tasks has been a cause of concern, especially in the case of organic farms that do not use chemicals. Combined with the knowledge that one third of all land in the mainland US is used for cattle grazing, this problem holds considerable significance. The lab has designed an autonomous battery-powered mobile robot, called Cowbot, for weed control in the rough and challenging environment on cow pastures. Cow pastures are usually open fields and there is large variation in weed population with geographic location and time of the year. He will then present their work on two interesting research questions: budget-aware weed detection using aerial imagery and online trajectory planning for the Cowbot to efficiently use weed detection information.

Traditionally, detection and planning have been addressed as separate problems that do not account for the range of operation of mobile robots. This separation leads to mobile robots either completing only a part of the operation or needing to refuel and resume operations. He will will present our work on weed detection from aerial imagery that accounts for the available planning budget of the autonomous mower. The second problem addresses online trajectory planning for the Cowbot with a limited field of view of onboard sensors and a finite turning radius. Given an onboard weed detection module, efficiently using detection information in real time to plan robot trajectories is challenging. Due to the unknown and variable weed density on pastures, coverage paths can lead to large wastage of resources. I will present reactive planning algorithms to compute efficient robot trajectories that utilize detection information from onboard sensing systems. They have deployed these algorithms on the Cowbot and have evaluated them in large scale experiments on cow pastures. He will then show videos of the Cowbot in action and talk about future directions that we are pursuing in this space.

Biography

Parikshit Maini is a Post-Doctoral Associate in the Department of Computer Science and Engineering at University of Minnesota and a member of the Robotic Sensor Networks lab headed by Prof. Volkan Isler. He works in the area of field robotics and applied AI with a focus on environmental and agricultural applications for mobile robot systems. He is leading the planning and navigation team on the "Cowbot - autonomous weed mower" project that has been covered in multiple news media stories (PBS, Star Tribune, Rural Media Group) and was recently showcased in live demos at the Minnesota FarmFest 2021. He also works on cooperative planning for heterogeneous multi-robot systems. He has developed planning algorithms for large-scale area coverage, persistent monitoring and visibility-based monitoring on terrains using cooperative aerial and ground robotic sensor nodes.

He holds a PhD in Computer Science and Engineering from Indraprastha Institute of Information Technology-Delhi, India. He also holds a M.Tech. degree in Computer Science and Engineering from IIIT-Delhi and a B.Tech. in Information Technology from Guru Gobind Singh Indraprastha University, Delhi in India.

The Minnesota Natural Language Processing (NLP) Seminar is a venue for faculty, postdocs, students, and anyone else interested in theoretical, computational, and human-centric aspects of natural language processing to exchange ideas and foster collaboration. The talks are every other Friday from 12 p.m. - 1 p.m. during the Fall 2021 semester.

This week's speaker, Dheeraj Rajagopal (Carnegie Mellon University), will be giving a talk titled "Explanations in Natural Language: From Theory to Practice."

Abstract

Understanding the reasoning process through explanations is spontaneous, ubiquitous and fundamental to our sense of perceiving the world around us. Scientific progress often relies on explanations to facilitate discovery of hypotheses, identify applications and also identify systematic errors and correct them. An in-depth study of explanations thus helps shed light on core cognitive issues, such as learning, induction and conceptual representation. Current NLP systems, despite significant advances, are usually treated as black boxes with little to no insight into how they reason. Understanding Explanations is an under-explored area in the natural language processing literature due to the lack of a unified formalism. In this talk, we will discuss two kinds of approaches to explanation - data-based and model-based. Next, we present methods for learning explanations via data (both statically and dynamically). We will also discuss an instantiation of model-based explanation approach where we explore how to explain model predictions while also jointly optimizing for the end-task. Finally, we discuss the learnings from the work and how to make the bridge between theory and practice for explanations in NLP.

Biography

Dheeraj Rajagopal is a 5th year PhD student at Carnegie Mellon University, School of Computing Sciences and advised by Prof. Eduard Hovy. His research interests are in the area of explanations, and designing explainable systems for Natural Language Processing. He has been an intern previously at MSR with the Knowledge Technologies and Experience team and at Ai2 with the Aristo team. Previously, he was a masters student at CMU working with Prof. Eduard Hovy and Prof. Teruko Mitamura.