Posts by Collection

Preface: I have never sat on an admissions committee, and my thoughts are based on conversations I have been able to observe, mostly on Twitter. These aren’t necessarily representative of my department, and I can’t speak to any specific university. This is a compilation of overarching themes I’ve observed from a variety of (primarily USA-based) academics. My thoughts are obviously going to be skewed by who I do (and, more importantly, don’t) know.

Diversity and Inclusion. It’s almost as much of a buzzword in Comp Sci as “Machine Learning” right now. Personally, I tend to lump the two together without parsing them as two separate concepts most of the time, so I’d like to take a second to break the two up.

Here is my research statement that I submitted for the NSF GRFP. Personal statements and reviews are available upon request as well.

The following is an assortment of young researchers I’ve had the pleasure of interacting with, either personally or by reading their work. If you find my work interesting, you might want to check them out as well. I’ll try to update it as I meet more people in my experience.

======

Short description of portfolio item number 1

Short description of portfolio item number 2

Abstract: Egocentric, or first-person, vision which became popular in recent years with an emerge in wearable technology,is different than exocentric (third-person) vision in some distinguishable ways, one of which being that the camera-wearer is generally not visible in the video frames. Recent work has been done on action and object recognition in egocentric videos, as well as work on biometric extraction from first-person videos. Height estimation can be a useful feature for both soft-biometrics and object tracking. Here, we propose a method of estimating the height of an egocentric camera without any calibration or reference points. We used both traditional computer vision approaches and deep learning in order to determine the visual cues that results in best height estimation. Here, we introduce a framework inspired by two stream networks comprising of two Convolutional Neural Networks, one based on spatial information, and one based on information given by optical flow in a frame. Given an egocentric video as an input to the framework, our model yields a height estimate as an output. We also incorporate late fusion to learn a combination of temporal and spatial cues. Comparing our model with other methods we used as baselines, we achieve height estimates for videos with a Mean Average Error of 14.04 cm over a range of 103 cm of data, and classification accuracy for relative height (tall, medium or short) up to 93.75% where chance level is 33%.

Abstract: In ethics, many academics make the assumption that all people want to be good. Evil comes in where there is a conflict of good decisions; where a decision that is good for one person contradicts the good of another. In this case, a person will make a different decision depending on their definition of the good they want to accomplish. In a society that starts with an equal proportion of selfishly good and selfessly good people, we aim to investigate the convergence of behavior through simulating the Iterated Prisoner’s Dilemma over time.

Download paper here

Abstract: In this paper, we utilize a multi-objective genetic algorithm (GA) to investigate the Iterated Prisoner’s Dilemma problem with a population of players that don’t have uniform objectives. Each of the members of our population has one of four objective pairs. We simulate a tournament similar to those in previous work to investigate patterns of convergence in objective pairs when they are free to change. We also consider the most successful objective pair within a population when members’ objective pairs are fixed.

A property or statistic of a distribution is said to be elicitable if it can be expressed as the minimizer of some loss function in expectation. Recent work shows that continuous real-valued properties are elicitable if and only if they are identifiable, meaning the set of distributions with the same property value can be described by linear constraints. From a practical standpoint, one may ask for which such properties do there exist convex loss functions. In this paper, in a finite-outcome setting, we show that in fact every elicitable real-valued property can be elicited by a convex loss function. Our proof is constructive, and leads to convex loss functions for new properties.

Abstract: We formalize and study the natural approach of designing convex surrogate loss functions via embeddings, for problems such as classification, ranking, or structured prediction. In this approach, one embeds each of the finitely many predictions (e.g. rankings) as a point in Rd, assigns the original loss values to these points, and “convexifies” the loss in some way to obtain a surrogate. We establish a strong connection between this approach and polyhedral (piecewise-linear convex) surrogate losses: every discrete loss is embedded by some polyhedral loss, and every polyhedral loss embeds some discrete loss. Moreover, an embedding gives rise to a consistent link function as well as linear surrogate regret bounds. Our results are constructive, as we illustrate with several examples. In particular, our framework gives succinct proofs of consistency or inconsistency for various polyhedral surrogates in the literature, and for inconsistent surrogates, it further reveals the discrete losses for which these surrogates are consistent. We go on to show additional structure of embeddings, such as the equivalence of embedding and matching Bayes risks, and the equivalence of various notions of non-redudancy. Using these results, we establish that indirect elicitation, a necessary condition for consistency, is also sufficient when working with polyhedral surrogates.

Abstract: A common technique in supervised learning with discrete losses, such as 0-1 loss, is to optimize a convex surrogate loss over R^d, calibrated with respect to the original loss. In particular, recent work has investigated embedding the original predictions (e.g. labels) as points in R^d, showing an equivalence to using polyhedral surrogates. In this work, we study the notion of the embedding dimension of a given discrete loss: the minimum dimension d such that an embedding exists. We characterize d-embeddability for all d, with a particularly tight characterization for d=1 (embedding into the real line), and useful necessary conditions for d>1 in the form of a quadratic feasibility program. We illustrate our results with novel lower bounds for abstain loss.

Abstract: The Iterated Prisoner’s Dilemma (IPD) has been studied in fields as diverse as economics, computer science, psychology, politics, and environmental studies. This is due, in part, to the intriguing property that its Nash Equilibrium is not globally optimal. Typically treated as a single-objective problem, a player’s goal is to maximize their own score. In some work, minimizing the opponent’s score is an additional objective. Here, we explore the role of explicitly optimizing for mutual cooperation in IPD player performance. We implement a genetic algorithm in which each member of the population evolves using one of four multi-objective fitness functions: selfish, communal, cooperative, and selfless, the last three of which use a cooperative metric as an objective. As a control, we also consider two single-objective fitness functions. We explore the role of representation in evolving cooperation by implementing four representations for evolving players. Finally, we evaluate the effect of noise on the evolution of cooperative behaviors. Testing our evolved players in tournaments in which a player’s own score is the sole metric, we find that players evolved with mutual cooperation as an objective are very competitive. Thus, learning to play nicely with others is a successful strategy for maximizing personal reward.

Abstract: As fairness and discrimination concerns permeate the design of both machine learning algorithms and mechanism design problems, we discuss differences in approaches between these two fields. We aim to bridge these two communities into a cohesive narrative that encompasses both the large-scale capabilities of machine learning and group-focused fairness as well as the strategic incentives and utility-based notions of fairness from mechanism de-sign, showing their necessity in designing a fair pipeline.

Abstract: In the field of design theory, the most well-known design is a Steiner Triple System. In general, a G-design on H is an edge-disjoint decomposition of H into isomorphic copies of G. In a Steiner Triple system, a complete graph is decomposed into triangles. In this paper we let H be a complete graph with ahole and G be a complete graph on four vertices minus one edge, also referred to as a K_4-e. A complete graph with a hole, K_d + v, consists of acomplete graph ond vertices,K_d, and a set of independent vertices of size v, V, where each vertex in V is adjacent to each vertex in K_d. Whend is even,we give two constructions for the decomposition of a complete graph with ahole into copies of K_4 -e: the Alpha-Delta Construction, and the Alpha-Beta-Delta Construction. By restricting d and v so that v = 2(d-1) - 5a, we are able to resolve both of these cases for a subset of K_d + v using difference methods and 1-factors.

Abstract: Given a prediction task, understanding when one can and cannot design a consistent convex surrogate loss, particularly a low-dimensional one, is an important and active area of machine learning research. The prediction task may be given as a target loss, as in classification and structured prediction, or simply as a (conditional) statistic of the data, as in risk measure estimation. These two scenarios typically involve different techniques for designing and analyzing surrogate losses. We unify these settings using tools from property elicitation, and give a general lower bound on prediction dimension. Our lower bound tightens existing results in the case of discrete predictions, showing that previous calibration-based bounds can largely be recovered via property elicitation. For continuous estimation, our lower bound resolves on open problem on estimating measures of risk and uncertainty.

Abstract: The Lovász hinge is a convex surrogate recently proposed for structured binary classification, in which k binary predictions are made simultaneously and the error is judged by a submodular set function. Despite its wide usage in image segmentation and related problems, its consistency has remained open. We resolve this open question, showing that the Lovász hinge is inconsistent for its desired target unless the set function is modular. Leveraging a recent embedding framework, we instead derive the target loss for which the Lovász hinge is consistent. This target, which we call the structured abstain problem, allows one to abstain on any subset of the k predictions. We derive two link functions, each of which are consistent for all submodular set functions simultaneously.

Abstract: Top-k classification is a generalization of multiclass classification used widely in information retrieval, image classification, and other extreme classification settings. Several hinge-like (piecewise-linear) surrogates have been proposed for the problem, yet all are either non-convex or inconsistent. For the proposed hinge-like surrogates that are convex (i.e., polyhedral), we apply the embedding framework Finocchiaro et al. (2019) to determine the prediction problem for which the surrogate is consistent. These problems can all be interpreted as variants of top-k classification, which may be better aligned with some applications. We leverage this analysis to derive constraints on the conditional label distributions under which these proposed surrogates become consistent for top-k. It has been further suggested that every convex hinge-like surrogate must be inconsistent for top-k. Yet, we use the same embedding framework to give the first consistent polyhedral surrogate for this problem.

Abstract: In the wake of increasing political extremism, online platforms have been criticized for contributing to polarization. One line of criticism has focused on echo chambers and the recommended content served to users by these platforms. In this work, we introduce the fair exposure problem: given limited intervention power of the platform, the goal is to enforce balance in the spread of content (e.g., news articles) among two groups of users through constraints similar to those imposed by the Fairness Doctrine in the United States in the past. Groups are characterized by different affiliations (e.g., political views) and have different preferences for content. We develop a stylized framework that models intra- and inter-group content propagation under homophily, and we formulate the platform’s decision as an optimization problem that aims at maximizing user engagement, potentially under fairness constraints. Our main notion of fairness requires that each group see a mixture of their preferred and non-preferred content, encouraging information diversity. Promoting such information diversity is often viewed as desirable and a potential means for breaking out of harmful echo chambers. We study the solutions to both the fairness-agnostic and fairness-aware problems. We prove that a fairness-agnostic approach inevitably leads to group-homogeneous targeting by the platform. This is only partially mitigated by imposing fairness constraints: we show that there exist optimal fairness-aware solutions which target one group with different types of content and the other group with only one type that is not necessarily the group’s most preferred. Finally, using simulations with real-world data, we study the system dynamics and quantify the price of fairness.

5 minute lightning talk at the 2018 Workshop on Aggregation, Dynamics, and Elicitation (WADE).

5 minute spotlight talk at the Neural Information Processing Systems 2018.

The slides from the presentation of my preliminary Ph.D. exam.

Poster presentation in Conference Proceedings at Neural Information Processing Systems 2019.

Poster presentation for the ICS annual poster session.

A talk summary and full talk are available here

The slides from the workshop talk.

Slides.

The slides from the presentation of my dissertation proposal (Qualifying exam).

Tutorial given with Manish Raghavan, Faidra Monachou, and Edwin Lock at EC 2021.

Link to talk (must be logged in to NeurIPS site) Slides Poster

Teaching Assistant for Graduate Algorithms in Spring 2019.

Introduction

Hi, I’m Jessie. My pronouns are she/her and I’m starting my 3rd year in the CS Theory group.

The deep dive

For class this week, we were asked to watched a video on the IDEO deep dive.

This week, I got to work with a handful of people, including (but not limited to… I didn’t catch all the names) Vinitha, Lucas, Isabella, and Adam.

Users and tasks

Formative research

Google VR headset personas

Interviews

Cultural probe design practice

For our project, we are interested in designing something to make sorting waste easier and more accurate. One major thing we need to understand with this is how people interact with their waste.

For a cultural probe related to our project, we can give volunteers recycling and compost bins to use in their house, observing where they place these bins (in relation to the landfill bin they already presumably have) and noting what waste items, if any, make their way into the recycling and compost bins.

Sketching and storyboarding

Interview practice

For this activity, Lucy interviewed me on activities I like to do. Being me, I talked about playing soccer and running, and most of her questions were directed at understanding how often I played and what the culture of my team was like. She asked how long I had played soccer and directed and adjusted her questions towards understanding the context in which I played soccer in order to understand why I love it.

Sketching practice

The first practice sketch we had was for an App to share your location with friends.

The second object we practiced sketching was another version of the good ole’ Apple TV remote from day 1.

Prototyping

Project storyboard practice

Here is a practice storyboard of someone interacting with our project design.

Study buddy prototype

The study buddy app paper prototype we did in class.

Usability

Visual Design

Usability heuristics: email inbox

In class, we talked about what some common heuristics for an email inbox were. Among many things, we discussed the organization of mail into folders, usuall managed on the left side. In order to start writing a new email, there is usually a bigger “compose” button that is differentiated by color.

Moreover, once you’re typing the email, the design of the email box is such that one works from top to bottom, starting with addressing the letter, having a greeting or subject line, then proceeding with the body of the message. Once all that is done, you can send the email by hitting the send button below all these entries.

Gesture design

In class, we discussed what makes gestures useful and easy to execute, and were tasked with designing our own gesture for a task. In cars, my table discussed gestures that could be associated with driving. My favorite alternative gesture was flicking someone off could be an alternative way to blare your horn at them. Alternatively, we could have taps on the back of the steering wheel with the right hand skip forward to the next song, if applicable, so that driver hands do not need to leave the wheel to navigate music.

KLM Model

In class, we discussed the KLM (Keystroke-level model) way of estimating performance of different interfaces depending on the number of clicks, points, mental preparation, drawing points, and hand relocations required. In class, I faced off against Chance to complete a form on Shaun’s test code and got destroyed. I think he took 14 seconds, where I took 21 seconds to fill out the form, which was closer to what our classmates calculated to be an expected completion time.