The Women in Machine Learning (WiML) workshop was founded in 2006 to forge connections within the relatively small community of women working in machine learning, to encourage mentorship, exchange of ideas, and promote communication. The program features 4 invited talks, 4 breakout sessions each having 2-8 parallel webinars, a panel with discussions on “industry/academic research, how to choose your path” and “post-pandemic adjustment and tips”, a mentoring social and 4 sponsor expo talks. Please refer to https://wimlworkshop.org/icml2021/program/ for more information.
The workshop attracts representatives from both academia and industry, whose contributed talks showcase some of the cutting-edge research done by women. In addition to technical presentations and discussion, the workshop aims to incite debate on promising research avenues and career choices for machine learning professionals. Details about WiML’s history and past events can be found at www.wimlworkshop.org. WiML workshops are overseen by the WiML Board of Directors, who select and oversee the organizing committee for each year’s workshop.
Wed 6:40 a.m. - 6:50 a.m.
|
Introduction & Opening Remarks
(Intro)
SlidesLive Video » |
Wenshuo Guo 🔗 |
Wed 6:50 a.m. - 7:00 a.m.
|
Diversity & Inclusion talk by D&I chair
(Talk)
SlidesLive Video » |
Danielle Belgrave 🔗 |
Wed 7:00 a.m. - 7:25 a.m.
|
Invited Talk #1 - Evaluating approximate inference for BNNs
(Talk)
SlidesLive Video » Bayesian Neural Network is one of the major approaches for obtaining uncertainty estimates for deep learning models. Key to the success is the selection of the approximate inference algorithms used to compute the approximate posterior, with mean-field variational inference (MFVI) and MC-dropout being the most popular variants. But is the good downstream uncertainty estimation performance of BNNs attributed to good approximate inference? In this talk I will discuss some of our recent results towards answer this question. I will also discuss briefly the computational reasons of the preference of MFVI/MC-dropout and describe our latest work to make BNNs more memory efficient. |
Yingzhen Li 🔗 |
Wed 7:00 a.m. - 7:15 a.m.
|
QuantumBlack Booth (Sponsor Booth) link » | 🔗 |
Wed 7:25 a.m. - 8:30 a.m.
|
Breakout Session 1.8: Neural Machine Translation for Low-Resource Languages
(Breakout Session)
link »
Passcode: 667130 Abstract: Neural Machine Translation (NMT) has seen a tremendous spurt of growth in less than ten years, and has already entered a mature phase. While considered as the most widely used solution for Machine Translation, its performance on low-resource language pairs still remains sub-optimal compared to the high-resource counterparts, due to the unavailability of large parallel corpora. Therefore, the implementation of NMT techniques for low-resource language pairs has been receiving the spotlight in the recent NMT research arena, thus leading to a substantial amount of research reported on this topic. This paper presents a detailed survey of research advancements in low-resource language NMT (LRL-NMT), along with a quantitative analysis aimed at identifying the most popular solutions. Based on our findings from reviewing previous work, this survey paper provides a set of guidelines to select the possible NMT technique for a given LRL data setting. It also presents a holistic view of the LRL-NMT research landscape and provides a list of recommendations to further enhance the research efforts on LRL-NMT. |
🔗 |
Wed 7:25 a.m. - 8:30 a.m.
|
Breakout Session 1.4: Unsupervised Learning in Computer Vision
(Breakout Session)
link »
In this session, we will introduce some of the most relevant computer vision problems, and explore them through the lens of unsupervised learning to highlight when and how it can be beneficial. |
🔗 |
Wed 7:25 a.m. - 8:30 a.m.
|
Breakout Session 1.3: Data Integration and Predictive Modeling for Precision Medicine in Oncology
(Breakout Session)
link »
Precision medicine in oncology aims to find a targeted treatment for each cancer patient sub-type or individual patient. This session will discuss how leveraging data-driven machine learning methods for integrating high-dimensional multi-omics cancer datasets and modelling their interplay can lead to improved risk assessment, early detection, and tailored cancer therapies for cancer patients. |
🔗 |
Wed 7:25 a.m. - 8:30 a.m.
|
Breakout Session 1.2: School mapping using computer vision technology (Breakout Session) link » | 🔗 |
Wed 7:25 a.m. - 8:30 a.m.
|
Breakout Session 1.1: Catching Out-of-Context Misinformation with Self-supervised Learning
(Breakout Session)
link »
Despite the recent attention to DeepFakes and other forms of image manipulations, one of the most prevalent ways to mislead audiences on social media is the use of unaltered images in a new but false context, commonly known as out-of-context image use. The danger of out-of-context images is that little technical expertise is required, as one can simply take an image from a different event and create a highly convincing but potentially misleading message. At the same time, it is extremely challenging to detect misinformation based on out-of-context images given that the visual content by itself is not manipulated; only the image-text combination creates misleading or false information. In order to detect these out-of-context images, several online fact-checking initiatives have been launched by newsrooms. However, they all heavily rely on manual human efforts to verify each post factually and to determine if a fact-checking claim should be labeled as "out-of-context". In this talk, I will discuss how can we build models that help determine the conflicting image-caption pairs and could be potential out-of-context misuse. |
🔗 |
Wed 7:25 a.m. - 8:30 a.m.
|
Breakout Session 1.6: Fundamentals of Contrastive Learning in Vision
(Breakout Session)
link »
In this breakout session, we will discuss the motivation behind contrastive learning and various recent contrastive models. We will discuss the intuition behind when and why contrastive learning works and how to improve its performance. Finally, we will discuss some tricks for successful application of contrastive learning in vision |
🔗 |
Wed 7:25 a.m. - 8:30 a.m.
|
Breakout Session 1.7: Exploring probabilistic sparse inferencing through the triangulation of neuroscience, computing and philosophy (Breakout Session) link » | 🔗 |
Wed 7:25 a.m. - 8:30 a.m.
|
Breakout Session 1.5: Machine Learning for Privacy: An Information Theoretic Perspective
(Breakout Session)
link »
A discussion on information leakage in machine learning models and concerns in everyday life |
🔗 |
Wed 7:30 a.m. - 9:00 a.m.
|
Salesforce Booth (Sponsor Booth) link » | 🔗 |
Wed 8:30 a.m. - 9:00 a.m.
|
Coffee Break and Posters AM (Poster Session) link » | 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Machine Learning Applications in Animal Sciences (Poster Session) link » | Ambreen Hamadani 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Emulating Aerosol Microphysics with Machine Learning (Poster Session) link » | Paula Harder 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Network Experiment Design for estimating Direct Treatment Effects (Poster Session) link » | Zahra Fatemi 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Adversarial Robust Model Compression using In-Train Pruning (Poster Session) link » | Sreetama Sarkar 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Iterative symbolic regression for learning transport equations
(Poster Session)
|
Heta Gandhi 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Cost Aware Asynchronous Multi-Agent Active Search (Poster Session) link » | ARUNDHATI BANERJEE 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Exploration and preference satisfaction trade-off in reward-free learning (Poster Session) link » | Noor Sajid 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
HYBRIDNET: A NETWORK THAT LEVERAGES ON CLASSICAL AND NON-CLASSICAL COMPUTER VISION TECHNIQUES FOR FEW SHOT LEARNING ON INFRARED IMAGERY (Poster Session) link » | Maliha Arif 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Clustering With Financial Fundamentals (Poster Session) link » | Jennifer Glenski 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Application of Knowledge Graph in Industry (Poster Session) link » | Samira Korani 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Contrastive Domain Adaptation (Poster Session) link » | Mamatha Thota 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Risk Analytics for Renewal of Purchase Orders (Poster Session) link » | Shubhi Asthana 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
On the (Un-)Avoidability of Adversarial Examples (Poster Session) link » | Sadia Chowdhury 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Extraction of Adverse Drug Reactions from Tweets using Aspect Based Sentiment Analysis (Poster Session) link » | Sukannya Purkayastha 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Interpretation and transparency in acoustic emotion recognition (Poster Session) link » | Sneha Das 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Seasonal forecasts of New Zealand's local climate conditions with limited GCM inputs using Convolutional Neural Networks (Poster Session) link » | Fareeda Begum 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Assessing the Carbon Intensity of Models Across Different Languages (Poster Session) link » | Krithika Ramesh 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
A Low-rank Support Tensor Network (Poster Session) link » | Kirandeep Kour 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
CricNet : Segment and Classify Cricket Events (Poster Session) link » | Shambhavi Mishra 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Episodically optimized dynamical networks for robust motor control (Poster Session) link » | Sruti Mallik 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Open Set Detection via Similarity Learning (Poster Session) link » | Sepideh Esmaeilpour 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
A modified limited memory Nesterov’s accelerated quasi-Newton (Poster Session) link » | Indrapriyadarsini Sendilkkumaar 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Time-series Forecasting of Ionospheric Space Weather using Ensemble Machine Learning (Poster Session) link » | Randa Natras 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
SocialBERT : An Effective Few Shot Learning Framework Applied to a Social TV Setting (Poster Session) link » | Debarati Das 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Explainable Prediction of Text Complexity: The Missing Preliminaries for Text Simplification (Poster Session) link » | Cristina Garbacea 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Alignment of Language Agents in Video games (Poster Session) link » | Gema Parreno 🔗 |
Wed 8:30 a.m. - 6:00 p.m.
|
Using Weak Supervision to Identify Drug Mentions from Class Imbalanced Twitter Data (Poster Session) link » | Ramya Tekumalla 🔗 |
Wed 8:30 a.m. - 9:00 a.m.
|
DeepMind Booth (Sponsor Booth) link » | 🔗 |
Wed 9:00 a.m. - 9:25 a.m.
|
Invited Talk #2 - Towards fairness & robustness in machine learning for dermatology
(Talk)
SlidesLive Video » Recent years have seen an overwhelming body of work on fairness and robustness in Machine Learning (ML) models. This is not unexpected, as it is an increasingly important concern as ML models are used to support decision-making in high-stakes applications such as mortgage lending, hiring, and diagnosis in healthcare. Currently, most ML models assume ideal conditions and rely on the assumption that test/clinical data comes from the same distribution of the training samples. However, this assumption is not satisfied in most real-world applications; in a clinical setting, we can find different hardware devices, diverse patient populations, or samples from unknown medical conditions. On the other hand, we need to assess potential disparities in outcomes that can be translated and deepen in our ML solutions. In this presentation, we will discuss how to evaluate skin-tone representation in ML solutions for dermatology and how we can enhance the existing models’ robustness by detecting out-out-distribution test samples (e.g., new clinical protocols or unknown disease types) over off-the-shelf ML models. |
Celia Cintas 🔗 |
Wed 9:25 a.m. - 10:30 a.m.
|
Breakout Session 2.3: Challenges and Opportunities in ML for Health Care: How to address interpretability in clinical decision making?
(Breakout Session)
link »
This un-workshop breakout session will discuss challenges and opportunities in ML for Health Care with a focus on interpretability in clinical decision making, the automation of clinical tasks, and fairness and bias concerns. |
🔗 |
Wed 9:25 a.m. - 10:30 a.m.
|
Breakout Session 2.7: Explainable machine learning: do we have the right tools?
(Breakout Session)
link »
We will discuss the present tools, challenges & future of explainability research. |
🔗 |
Wed 9:25 a.m. - 10:30 a.m.
|
Breakout Session 2.1: Geometry and Machine Learning (Breakout Session) link » | 🔗 |
Wed 9:25 a.m. - 10:30 a.m.
|
Breakout Session 2.8: Decision-Making in Social Settings: Addressing Strategic Feedback Effects
(Breakout Session)
link »
Consequential decisions compel individuals to strategically react in response to the specifics of the decision rule. In aggregate, these individual-level responses can disrupt the statistical patterns that motivated the decision rule, leading to unforeseen consequences. In this breakout session, we will give an introduction to the area of strategic classification, with a focus on limitations of the standard models of agent behavior. We will then have a discussion about applications where strategic adaptation is observed, potential approaches for modeling strategic agents, and connections to other domains. |
🔗 |
Wed 9:25 a.m. - 10:30 a.m.
|
Breakout Session 2.2: Leveraging Open-Source Tools for Natural Language Processing
(Breakout Session)
link »
Natural language processing (NLP), or the task of turning human language into structured data for machines to understand, powers the chatbots, voice assistants, predictive text, and other pervasive language applications that make our lives easier. We feel open-source tools complement the rapidly evolving field of NLP by helping to drive innovation through collaboration. In this breakout session, participants will be invited to share their experiences and lessons learned using open-source tools for NLP and explore how open-source tools may be able to help with current challenges in NLP. |
🔗 |
Wed 9:25 a.m. - 10:30 a.m.
|
Breakout Session 2.4: Leading the Way for the Next Generation of Black Women in STEM
(Breakout Session)
link »
This session will provide an opportunity to explore tools and techniques to assist the next generation of Black women as they pursue STEM careers and STEM studies. |
🔗 |
Wed 9:25 a.m. - 10:30 a.m.
|
Breakout Session 2.5: Un-bookclub Algorithms of Oppression (Breakout Session) link » | 🔗 |
Wed 9:25 a.m. - 10:30 a.m.
|
Breakout Session 2.6: Research within community: how to cultivate a nurturing environment for your research
(Breakout Session)
link »
Even though it is rarely emphasized, having a supportive environment is crucial to one’s research success. While in a lot of cases we don’t really have 100% control over what lab environment, what peers, what advisors we end up finding ourselves in, there are still things we can do to cultivate an environment, a community, even a culture, that works for us. |
🔗 |
Wed 9:30 a.m. - 10:30 a.m.
|
Salesforce Booth (Sponsor Booth) link » | 🔗 |
Wed 10:30 a.m. - 10:45 a.m.
|
Microsoft: Improving productivity with Graph ML over content-interaction networks
(Expo Talk)
SlidesLive Video » |
Jennifer Neville 🔗 |
Wed 10:45 a.m. - 11:00 a.m.
|
QuantumBlack: Algorithmic Fairness (Machine Learning with a Human Face)
(Expo Talk)
SlidesLive Video » |
Viktoriia Oliinyk 🔗 |
Wed 10:45 a.m. - 11:45 a.m.
|
Facebook Booth (Sponsor Booth) link » | 🔗 |
Wed 10:45 a.m. - 11:00 a.m.
|
QuantumBlack Booth (Sponsor Booth) link » | 🔗 |
Wed 11:00 a.m. - 11:15 a.m.
|
Apple: Machine Learning at Apple
(Expo Talk)
SlidesLive Video » |
Lizi Ottens 🔗 |
Wed 11:15 a.m. - 11:30 a.m.
|
Facebook: Future of AI-Powered Shopping
(Expo Talk)
SlidesLive Video » |
Ning Zhang 🔗 |
Wed 11:30 a.m. - 12:30 p.m.
|
Mentoring Room with Anna Goldenberg, University of Toronto - Two body problem in academia, raising a family, grant strategies, looking for a job, and deploying ML in a hospital setting (Mentoring) link » | 🔗 |
Wed 11:30 a.m. - 12:30 p.m.
|
Mentoring Room with Lalana Kagal from MIT (Mentoring) link » | 🔗 |
Wed 11:30 a.m. - 12:30 p.m.
|
Mentoring Room with Dina Obeid from Harvard - A non-linear career path in machine learning (Mentoring) link » | 🔗 |
Wed 11:30 a.m. - 12:30 p.m.
|
Mentoring Room with Been Kim from Google Brain - Industry research and managing up (Mentoring) link » | 🔗 |
Wed 11:30 a.m. - 12:30 p.m.
|
Mentoring Room with Shakir Mohamed from DeepMind - Socio-Technical AI Research (Mentoring) link » | 🔗 |
Wed 11:30 a.m. - 12:30 p.m.
|
Mentoring Room with Angelique Taylor from UC San Diego - Transitioning from PhD to Assistant Professor (Mentoring) link » | 🔗 |
Wed 12:30 p.m. - 1:30 p.m.
|
Facebook Booth (Sponsor Booth) link » | 🔗 |
Wed 2:00 p.m. - 2:30 p.m.
|
Microsoft Booth (Sponsor Booth) link » | 🔗 |
Wed 3:45 p.m. - 4:25 p.m.
|
Invited Talk #3 - Characterizing the Generalization Trade-offs Incurred By Compression
(Talk)
SlidesLive Video » To-date, a discussion around the relative merits of different compression methods has centered on the trade-off between level of compression and top-line metrics such as top-1 and top-5 accuracy. Along this dimension, compression techniques such as pruning and quantization are remarkably successful. It is possible to prune or heavily quantize with negligible decreases to test-set accuracy. However, top-line metrics obscure critical differences in generalization between compressed and non-compressed networks. In this talk, we will go beyond test-set accuracy and discuss some of my recent work measuring the trade-offs between compression, robustness and algorithmic bias. Characterizing these trade-offs provide insight into how capacity is used in deep neural networks — the majority of parameters are used to represent a small fraction of the training set. Formal auditing tools like Compression Identified Exemplars (CIE) also catalyze progress in training models that mitigate some of the trade-offs incurred by compression. |
Sara Hooker 🔗 |
Wed 4:25 p.m. - 5:30 p.m.
|
Breakout Session 3.1: Does your model know what it doesn’t know? Uncertainty estimation and out-of-distribution (OOD) detection in deep learning (Breakout Session) link » | 🔗 |
Wed 4:25 p.m. - 5:30 p.m.
|
Breakout Session 3.5: Variational Inference for Neural Networks (Breakout Session) link » | 🔗 |
Wed 4:25 p.m. - 5:30 p.m.
|
Breakout Session 3.2: ML Applications in Big Code
(Breakout Session)
link »
With the emergence of publicly shareable repositories on the Web (i.e. open sourced code in GitHub), there are now billions of lines of code easily to read and learn from. With the abundance of available code artifacts, researchers have begun to apply machine learning techniques to find statistical patterns in Big Code to guide developers to understand and write code better. This breakout session will cover the applications of Big Code, including program synthesis, code search, code autocomplete, code bug fixing. |
🔗 |
Wed 4:25 p.m. - 5:30 p.m.
|
Breakout Session 3.3: Connecting Novel Perspectives on GNNs: A Cross-Domain Overview
(Breakout Session)
link »
Humans represent complex systems as compositions of entities and interactions among these entities, i.e., a graph. In this session, we particularly focus on the algorithmic and theoretical foundations of Graph Neural Networks (GNNs). GNNs are a family of methods that generalize and extend neural networks to operate on relational data, and provides a flexible interface for manipulating structured knowledge, learning structured representations, and relational reasoning. Despite the rapid growth in the past five years, there are several limitations to applying and generalizing current GNN techniques to model datasets in novel applications, e.g., computer vision, graphics, wireless communication, etc.” |
🔗 |
Wed 4:25 p.m. - 5:30 p.m.
|
Breakout Session 3.4: Bridging the gap between academia and industry (Breakout Session) link » | 🔗 |
Wed 4:25 p.m. - 5:30 p.m.
|
Breakout Session 3.6: Responsible AI in production: Technical and Ethical considerations (Breakout Session) link » | 🔗 |
Wed 4:30 p.m. - 5:30 p.m.
|
Salesforce Booth (Sponsor Booth) link » | 🔗 |
Wed 5:30 p.m. - 6:00 p.m.
|
Coffee Break and Posters PM (Poster Session) link » | 🔗 |
Wed 6:00 p.m. - 6:25 p.m.
|
Invited Talk #4 - Errors are a crucial part of dialogue
(Talk)
SlidesLive Video » Collaborative grounding is a fundamental aspect of human-human dialogue which allows people to negotiate meaning; in this talk, I argue that current deep learning approaches to dialogue systems don’t deal with it adequately. Making errors, and being able to recover from them collaboratively, is a key ingredient in grounding meaning, but current dialogue systems can’t do this. I will illustrate the pitfalls of being unable to ground collaboratively, discuss what can be learned from the language acquisition and dialog systems literature, and reflect on how to move forward. I will urge the community to proceed by addressing a research gap: how clarification mechanisms can be learned from data. Novel research methodologies which highlight the importance of the role of clarification mechanisms are needed for this. I will present an annotation methodology, based on a theoretical analysis of clarification requests, which unifies a number of previous accounts. Dialogue clarification mechanisms are an understudied research problem and a key missing piece in the giant jigsaw puzzle of natural language understanding. I will conclude this talk with an open call for collaborators that share the vision presented. |
🔗 |
Wed 6:25 p.m. - 7:30 p.m.
|
Breakout Session 4.1: AI and Creativity: Approaches to Generative Art (Breakout Session) link » | 🔗 |
Wed 6:25 p.m. - 7:30 p.m.
|
Breakout Session 4.2: Attrition of women and minoritized individuals in AI (Breakout Session) link » | 🔗 |
Wed 6:25 p.m. - 7:30 p.m.
|
Breakout Session 4.3: Safely navigating scalability-reliability trade-offs in ML methods
(Breakout Session)
link »
ML methods across the discipline make a variety of approximations in order to scale. We will discuss how to navigate these trade-offs appropriately, particularly in the context of high-impact ML applications. |
🔗 |
Wed 7:30 p.m. - 8:30 p.m.
|
Panel Discussion
SlidesLive Video » The first half of the panel will be focused on discussions on “industry/academic research, how to choose your path” and “post-pandemic adjustment and tips”. The second half will be an "ask me anything" session, where the panelists will be answering questions from the participants. Panelists - Sarah Aerni (Salesforce Einstein), Kalesha Bullard (Facebook), Sarah Dean (Cornell University), Sylvia Herbert (University of California, San Diego), and Amy Zhang (Facebook). |
🔗 |
Wed 8:30 p.m. - 8:45 p.m.
|
Closing Remarks
(Closing)
SlidesLive Video » |
Sarah Osentoski 🔗 |