Here is this semester's tentative schedule; I will update it as the semester progresses. All assignments will be due via the class Canvas, in accordance with SLU's policies.
All presentations are planned for 3:10-4pm in Kelley Auditorium, otherwise known as Lecture Halls 005, unless indicated otherwise in the table below. (Occasional events may take place on zoom for virtual speakers, but these will be announced if so.)
Date Title Speaker Abstract Bio August 17 Introduction to the course Erin Chambers January 25 Towards Intelligent Next-Generation Wireless Networks Nan Cen Rapid growth in the number of connected wireless devices (e.g., mobile phones, low-power IoT devices, connected vehicles, etc.) will expand the scale of the next-generation wireless and mobile networks. Moreover, the foreseen applications like digital twins, smart homes and cities, connected autonomous vehicles, Industry 4.0, and agriculture precision will require network intelligence and new spectrum technologies to enable ultra-low latency and highly reliable communication in complex 5G and beyond wireless networks. In this talk, we will discuss how to add intelligence in next-generation wireless networks, including: (1) new 6G spectrum technology, i.e., wireless visible light networking, to complement well with current crowded RF networks; (2) learning-based networking control algorithms design to optimize the system performance; (3) software-defined intelligent network architecture design to enable fast prototyping of NextG networking testbed. Dr. Nan Cen is an Assistant Professor in the Department of Computer Science at Missouri University of Science and Technology. She received her Ph.D. in Electrical and Computer Engineering from Northeastern University in 2019, under the supervision of Professor Tommaso Melodia. She received her M.S. in Electrical Engineering from SUNY Buffalo in 2014. She had previously received her B.S. and M.S. in Information Science and Engineering from Shandong University in 2008 and 2011. Dr. Cen’s research is focused on modeling, control, and optimization for next-generation intelligent wireless systems and networks, data-driven networking control algorithms design, software defined radios based testbed implementation, and Artificial Intelligent of Things. She has chaired N2Women workshop on WoWMoM 2022 and served TPC member for EAI WiCon 2022, IEEE INFOCOM Workshop WCNEE 2020 AND 2021, IEEE/ACM IWQoS 2020, etc. Dr. Cen is a member of IEEE and ACM. February 1 Lecture Video Analysis and Its Applications Kenny Mauricio Davila Castellanos The internet is full of educational videos. Many instructors record themselves explaining complex topics, and then publish these videos with the goal of helping students from different educational levels. Among these, we find a large number of videos of math-oriented lectures and tutorials covering topics that are better explained using handwritten content on whiteboards or chalkboards. In this talk, I will discuss methods for analyzing different aspects of such lecture videos including the handwritten content, the speaker actions and the temporal segments of the lecture. The information produced by these methods enables the creation of tools for advanced applications such as video summarization, navigation and search. In turn, these applications can help both students and instructors find useful educational resources when they need them. Dr. Kenny Mauricio Davila Castellanos completed his PhD in Computing and Information Sciences at Rochester Institute of Technology. His research overlaps multiple fields such as document analysis, computer vision, pattern recognition and information retrieval, with a recent focus on topics that include extraction, classification and recognition of charts as well as retrieval of math formulas from documents and videos. He completed a post-doctorate at the Center for Unified Biometrics and Sensors (CUBS) - University at Buffalo under the supervision of Venu Govindaraju and Srirangaraj Setlur, and worked at the Document and Pattern Recognition Lab (DPRL) under the supervision of Dr. Richard Zanibbi February 8 Toward Practical Program Repair Ali Ghanbari Automatic program repair (APR) is one of the recent advances in automated software engineering aiming for reducing the burden of debugging by suggesting patches that either directly fix the bugs or help the programmers during manual debugging. Despite the remarkable progress of APR in the last decade, state-of-the-art techniques suffer from problems in three areas of scalability, applicability, and effective patch correctness assessment, reducing practicality of APR. In this talk, I will describe the steps that we have taken toward realizing practical APR by proposing solutions to alleviate each of the above-mentioned problems. As for scalability and applicability, we introduce and evaluate Java Virtual Machine (JVM) bytecode-level patch generation and validation which allows (1) on-the-fly patch generation and validation and (2) uniform treatment of programs written in dozens of programming languages. This speed-up also allows our technique to explore more of repair search space and find more genuine fixes than state-of-the-art. We also introduce and evaluate two techniques for correctness assessment of automatically generated patches via both ranking and classification. Ali Ghanbari is a Postdoctoral Research Associate in the Department of Computer Science at Iowa State University. He received his PhD in Software Engineering from The University of Texas at Dallas in 2021. His research interests span the intersection of Software Engineering, Programming Languages, and Data Science. Currently, he is focused on fault localization and automated repair of deep neural networks. Ali is also working on modular decomposition of deep neural networks and studying the impact of modularity in hiding user requirements changes. His postdoctoral research is funded by the National Science Foundation under the prestigious CIFellows program. During his PhD, he worked on addressing three challenges in the construction of practical program repair systems. His PhD research won 3rd place in Student Research Competition in ASE 2019 and led to the publication several papers in top software engineering venues and a novel fault localization technique which is being used in a multi-national tech compan February 15 Looking for Graves in All the Wrong Places (and other Social Justice Applications of Machine Learning) Abby Stylianou (tentative) In recent years, computer vision and machine learning systems have made huge strides towards performing at or beyond human performance on a variety of visual recognition and understanding tasks -- image classification, segmentation, and generation among them. These sorts of algorithms have been deployed in tools like Google Image Search, Facebook facial detection and identification, and in self-driving cars. In this talk, we focus on the use of computer vision and machine learning tools for the social good, with an emphasis on two use cases: identifying the location of a lost grave of a young murder victim, and determining what hotels human trafficking victims are photographed in. The talk will discuss not only the specific algorithmic details behind the different applications, but also the ethical concerns of building such powerful visual tools. Dr. Abby Stylianou has been an Assistant Professor of Computer Science at Saint Louis University since Fall of 2019. She received her B.A. in Environmental Studies: Geoscience from Washington University in St. Louis. She was then hired as a Research Scientist by Dr. Robert Pless, in the Media and Machines Lab at WashU, where her work focused on tools for the calibration and validation of outdoor imagery. She then pursued an M.S. in Computer Science part time while working as a Research Scientist, before decided to pursue a Ph.D., focused on large scale image search approaches to combat human trafficking by recognizing the hotels that victims of human trafficking are photographed. She continued working on this project as a Postdoc at George Washington University, working with Dr. Pless, and funded by the National Institute of Justice, before beginning as a professor at SLU. February 22 The Shape of Data: An Introduction to Topological Data Analysis David Letscher Topological data analysis (TDA) is a filed at the intersection of computer science, mathematics and statistics that provides the shape and structure of complex data. TDA uses mathematical tools from algebraic topology to extract topological features of data sets, which can be used to uncover hidden patterns and relationships. In this talk, we will discuss two main topics. First, we cover the TDA pipeline from data representation, feature extraction using persistent homology, and statistical methods for evaluating the significance of results. The second part of the talk will be about persistence-based techniques can be used to simplify data. Dr. Letscher received a B.S. in Mathematics from Notre Dame and a Ph.D. in Mathematics from the University of Michigan. After doing postdoctoral work in UC-San Diego and Oklahoma State and visiting positions at the University of Melbourne and the American Institute of Mathematics, Dr. Letscher joined Saint Louis University in 2001. Dr. Letscher's primary research is in computational topology but has done work in 3-manifold topology, mathematics of sports, computer science education. His current research includes works on shape analysis and simplification, machine learning techniques in computational topology and AI algorithms for playing card games. March 1 Data Who? Kyle Sykes The industry is filled with many “data _____” job titles, so deciphering them all can be confusing. How many ways can you realistically work with data? Are some of these actually the same? Which ones are someone with a CS degree prepared for? After this talk you should have some idea of what the different data-related roles are in the industry. We'll cover a lot of the common “data” related job roles and attempt to make some sense of it all. Kyle Sykes graduated from SLU in 2016 with a Ph.D. in Mathematics. His research area was in the field of computational topology/geometry. He has worked as a Data Scientist in the defense and healthcare industries and did consulting as a Data Engineer. He currently is doing Data/Software Engineering at a startup called Avise. March 8 APECS: A Distributed Access Control Framework for Pervasive Edge Computing Services Reza Tourani Edge Computing is a new computing paradigm where applications operate at the network edge, providing low-latency services with augmented user and data privacy. A desirable goal for edge computing is pervasiveness, that is, enabling any capable and authorized entity at the edge to provide desired edge services–pervasive edge computing (PEC). However, efficient access control of users receiving services and edge servers handling user data, without sacrificing performance is a challenge. Current solutions, based on “always-on” authentication servers in the cloud, negate the latency benefits of services at the edge and also do not preserve user and data privacy. In this paper, we present APECS, an advanced access control framework for PEC, which allows legitimate users to utilize any available edge services without need for communication beyond the network edge. The APECS framework leverages multi-authority attribute-based encryption to create a federated authority, which delegates the authentication and authorization tasks to semi-trusted edge servers, thus eliminating the need for an “always-on” authentication server in the cloud. Additionally, APECS prevents access to encrypted content by unauthorized edge servers. We analyze and prove the security of APECS in the Universal Composability framework and provide experimental results on the GENI testbed to demonstrate the scalability and effectiveness of APECS. Reza Tourani is an assistant professor with the Computer Science department at Saint Louis University since 2018. He received his PhD in Computer Science from New Mexico State University in 2018. His research interests are the broader areas of security and privacy, with a focus on building secure by design protocols and resilient systems, ranging from cyber-physical system such as Internet Things to the Future Internet Architecture. As part of his research, Dr. Tourani works on pressing security issues in edge computing ecosystem with a particular emphasis on distributed access control, trusted and verifiable computing using secure hardware and trusted execution environments, and secure computation outsourcing. March 22 Mythbusting Open Source for Software Professionals Daniel Shown Source code for software is widely available online, but many questions remain for software professionals. Why do organizations choose to release their proprietary code or collaborate with their competitors? What opportunities and obligations exist when working with open source software? How can software professionals engage in the world of open source software effectively and responsibly? Many myths have developed online and many professionals jeopardize their careers by working under false assumptions. There are a handful of critical things anyone developing software should know about working with open source software, and good resources available to help answer more complex questions. Daniel Shown is an artist, technologist and symmathesist. As an artist by training and an alumnus of Saint Louis University (BA Studio Art) he works in a range of traditional and new mediums. With decades of professional experience in software development and operations he has maintained parallel careers as a technologist and artist. He currently serves as the Program Director for Open Source with SLU and as an adjunct instructor in Computer Science. As both an artist and a technologist he engages symmathesy, learning systems made of learning parts, as a theoretical lens and an evolving pragmatic toolset. March 29 Advanced Computational Techniques for Real-Time Biological Sequence Analyses Ted Ahn The exponential advances in sequencing technologies and informatics tools for generating and processing large biological data sets are promoting a paradigm shift in the way we approach in bioinformatics and biomedical problems. The vast amount of resulting sequencing data can help the diagnosis and precision medicine but poses a challenge: accurately characterizing a sample in a computationally feasible fashion, despite specimen diversity. In this talk, I will describe how to solve bio- and biomedical informatics research problems using advanced computational techniques for aiming real-time biological sequence analyses. First, I will present metagenomic sequence analysis by classifying metagenomic samples using microbiota information and machine learning techniques to identify and predict disease samples precisely. Second, I will introduce immune cell sequencing analysis. Sequencing T-cell receptors (TCR) can provide us with a robust resource for understanding whether an individual has been infected with a pathogen after the infection occurred. Even after the infection is eliminated, pathogen-specific immune cells and their receptor sequences are present at higher frequencies than prior to infection, and their increase in frequency prevents secondary infections. We developed a scalable deep neural network model to advance our capabilities of identifying infections from pathogen-specific TCR sequences in circulation over time with exceptional sensitivity. infections from pathogen-specific TCR sequences in circulation over time with exceptional sensitivity. Tae-Hyuk (Ted) Ahn (Ph.D.) is an Associate Professor in the Department of Computer Science at Saint Louis University (SLU). He is also a core faculty member in the graduate program of Bioinformatics and Computational Biology. His research interests include bioinformatics, biomedical informatics, high-performance computing, big data analytics, and machine & deep learning. Before joining SLU in 2015, he worked at Oak Ridge National Laboratory as a postdoctoral researcher. He received the Ph.D. degree in Computer Science from Virginia Tech in 2012, the M.S. degree in Electrical and Computer Engineering from Northwestern University in 2007, and the B.S. degree in Electrical Engineering from Yonsei University in 2000. From 2000 to 2004, he worked in the industry at Samsung SDS in South Korea. April 5 Exploring the mathematical shape of plants Erik J. Amézquita Shape is foundational to biology. Observing and documenting shape has fueled biological understanding as the shape of biomolecules, cells, tissues, and organisms arise from the effects of genetics, development, and the environment. To comprehensively quantify the vast morphological diversity in plant biology, we focus thus on Topological Data Analysis (TDA). TDA is an emerging mathematical discipline that uses principles from algebraic topology to comprehensively measure shape in a broad scope of different datasets. Using a function that relates the similarity of data points to each other, we can monitor the evolution of core mathematical shape features ---connected components, loops, and voids. This evolution, a shape signature, concisely summarizes large, complex datasets. As a proof of concept, we explore three applications of TDA in plant biology based on X-ray CT scan 3D reconstructions. First, we analyze the shape of tens of thousands of barley seeds. With the Euler Characteristic Transform, topological shape descriptors of the seeds are used to successfully characterize different barley varieties based solely on the grain morphology. Understanding grain shape can allow us to select better grains for an ever-changing climate. Second, we use directional statistics and persistence homology to model the shape and distribution of citrus and their oil glands. Quantifying these distributions can allow us to understand better citrus fruit development. Finally, we comprehensively measure the shape of walnut shells and kernel to then explore how shape nuances can play significant differences in a walnut crackability and domestication. Combining traditional and topological shape descriptors, we explore the relationship between shell morphology and the force required to crack it open. The vision of TDA, that data is shape and shape is data, will be relevant as biology transitions into a data-driven era where meaningful interpretation of large datasets is a limiting factor. Erik Amézquita (he/him) is a mathematician by training, a computer scientist by trade, and a plant biologist by collaboration. He is currently a PhD candidate in the Department of Computational Mathematics, Science, and Engineering at Michigan State University. His research focuses on the mathematical quantification of plant morphology. He is especially interested in drawing mathematical tools from Topological Data Analysis (TDA) to better understand the shape of 3D and 2D plant scans. Before coming to the US, Erik did his math degree at Universidad de Guanajuato, in Mexico, with extensive support from the Centro de Investigación en Matemáticas. He has also collaborated in projects to lower interdisciplinary, language, and geographical barriers, like the bilingual learn-to-code notebooks of Plants & Python, and social data analysis that brings attention to the legacy of colonialism in plant science research. Starting this coming summer, Erik will be a PFFFD Postdoctoral Fellow at the Division of Plant Science and Technology in the University of Missouri. April 12 Constrained Optimization of Elastic Task Adaptation in Real-Time Systems Marion Sudvarg Applications ranging from simultaneous localization and mapping (SLAM) to prompt detection of astrophysical transients are subject to exogenous timing constraints whose satisfaction may be met according to application-specific objective functions. Optimization within those constraints requires new models and analyses that can offer rigorous assurance while exploiting opportunities to improve performance. Furthermore, in resource-constrained and mixed criticality environments, dynamic application requirements and resource availability may force tasks to adapt execution state to maintain deadline guarantees. In this talk, I will discuss the design of new elastic scheduling models that provide principled and provable approaches to adaptation for recurrent real-time task systems. These frameworks can compress task utilization to maximize the expected computational and functional performance of the application within the bounds of timing constraints. Extensions to these models reduce pessimism in resource allocation, especially for parallel tasks, for which finding an optimal schedule is generally intractable. I apply and evaluate this approach in the context of the Advanced Particle-astrophysics Telescope (APT) and its Antarctic Demonstrator (ADAPT), and as future work consider other relevant applications such as ORB-SLAM. Marion Sudvarg is a PhD candidate studying computer science at Washington University in St. Louis, advised by Dr. Christopher Gill. He earned bachelor’s degrees in math and physics from Wash U. He then worked as a network administrator and later an information systems manager. Driven by a love for computing, he returned to Wash U for part-time master’s studies in computer science and attained a graduate certificate in data mining and machine learning. He left industry after six years to pursue a full-time research career, beginning with doctoral studies. Marion’s research interests are in developing robust, adaptable, and secure real-time computing systems, and he works with the APT collaboration to develop real-time data analysis algorithms and systems for multi-messenger high-energy astrophysics. He has published in multiple physics and computer science venues and was recognized with the best paper award at RTCSA 2022. He additionally enjoys teaching, having taught Wash U’s advanced undergraduate Operating Systems Organization and graduate-level Advanced Operating Systems courses. He recently led a significant restructuring of content for both courses and developed the majority of material and assignments now used in Advanced Operating Systems. April 26 Towards morphing geodesic graphs on the sphere Christian Howard Given two graph embeddings Γ_0 and Γ_1 on a surface, a well studied class of problems in computational topology revolve around the existence and construction of continuous transformations between the two embeddings that preserve the embedding property, referred to as morphs, and potentially other properties, e.g. convexity. In this talk, we focus on a surprisingly challenging variant: morphs between geodesic spherical embeddings. With applications in areas like graphics, it is surprising that unlike many other surfaces, there has yet to be given a provably correct and sufficiently general geodesic spherical morphing algorithm that requires only a polynomial number of morphing steps. In this talk, we first discuss the simpler related variant of planar morphing and draw connections to the spherical case while also identifying the new challenges. We then discuss theoretical progress towards efficient spherical morphs and discuss some new promising directions to resolve the more general case. Christian Howard is a theoretical computer science PhD student from UIUC working with advisor Jeff Erickson on problems in computational geometry and topology. Prior to his PhD, he worked as an aerospace engineer developing algorithms in the context of robotics, scientific computing, machine learning, and control systems. April 19 Seed composition estimation from standing crops: A multitemporal, multimodal and multisensory approach to estimate soybean seed composition using machine learning Dr. Vasit Sagan Global food security has increasingly been strained by climate change and geopolitical strife. The increasing population requires 70% more food to feed the humanity in next several decades. With limited arable land and shrinking water resources, innovations in technologies including remote sensing, computing, AI/ML, and crop breeding become vital to improving food security while advancing agricultural sustainability. We present how very high-resolution UAV and satellite data are used to predict crop yields and seed composition from in-season standing crops. We compare the results from satellite imaging with drone-based estimation in order to investigate the scalability of remote sensing observations. We show that advances in AI and computing coupled with remote sensing offer new insights for global food security improvements and the development of more environmentally resilient farming systems. Vasit Sagan is Acting Director of the Taylor Geospatial Institute (TGI), Associate Professor in the Department of Earth and Atmospheric Sciences at Saint Louis University (SLU), and Director of SLU’s Remote Sensing Lab. Sagan leads the TGI’s efforts to harness the power of partnership with industry, government agencies, and research entities to develop geospatial research addressing some of the world’s greatest challenges. Previously, Dr. Sagan served as the founding Director of Saint Louis University’s Geospatial Institute (GeoSLU) and directed SLU’s campuswide geospatial research and training programs. Under his leadership, GeoSLU secured over $10M in external grants and contracts and played an important role in the creation of the Taylor Geospatial Institute. Sagan’s research focuses on developing state-of-the-art computer vision technologies, AI/machine learning, and sensor/information fusion algorithms for studying food and water security, ecosystems, and social instability from local to global scales. He has been PI/Co-PI on over $40M in grant funding and has authored over 150 peer-reviewed journal publications, many of which have been recognized through best paper awards. He has served on NASA review panels and reviewed several NSF proposals and numerous journal papers. He has also advised and mentored numerous doctoral students, master’s students, and postdocs and served as a member of dozens of doctoral dissertation committees. He currently serves as a member of the National Geospatial Advisory Committee under the U.S. Department of the Interior, where he is chairing the committee to advance U.S. global geospatial excellence and innovation. He is also Associate Editor of the ISPRS Journal of Photogrammetry and Remote Sensing, Director of the MissouriView Consortium, and Primary Investigator for NGA-SLU CRADA since 2018. He was involved in NASA’s Air Quality Applications Science Team and NASA’s recent mission focused on studying the atmospheric composition, TEMPO (Tropospheric Emissions: Measurement of Pollution). He was the Vice President and President of ASPRS Heartland Region from 2013-2015.