Dalin (Darlene) Guo

Bio

I am a Machine Learning Engineer at Twitter Cortex, working on recommender system research.

Prior to that, I am a PhD candidate in Cognitive Science at UC San Diego working with Prof. Angela Yu. I received B.S. in Electrical Engineering from Beijing Institute of Technology in 2016, and master degree in Intelligent Systems, Robotics, and Control from UC San Diego in 2018.

I was a Software Engineer and Machine Learning intern at Facebook working on news feed recommendation during summer 2020. I was a Machine Learning Researcher intern at Twitter Cortex London office working on deep reinforcement learning for online recommendation in summer 2019.

Research Interest

My research interest lies in machine learning and cognitive science. For my PhD research, I use computational methods to understand reward-based human learning and decision-making under uncertainty. My PhD thesis is on modeling human behavior in multi-armed bandit tasks, where an agent chooses among novel options, trading-off between exploration and exploitation.

Publications

Revisiting the Role of Uncertainty-Driven Exploration in a (Perceived) Non-Stationary World. CogSci 2021.[pdf]
Dalin Guo, Angela J. Yu. (2021)

Deep Bayesian Bandits: Exploring in Online Personalized Recommendations. Recsys 2020.[pdf] [poster] [video]
Dalin Guo, Sofia Ira Ktena, Pranay Kumar Myana, Ferenc Huszár, Michael Kneier, Sourav Das, Wenzhe Shi, Alykhan Tejani. (2020)

Devaluation of Unchosen Options: A Bayesian Account of the Provenance and Maintenance of Overly Optimistic Expectations. CogSci 2020.[pdf]
Corey Yishan Zhou, Dalin Guo, Angela J. Yu. (2020)

A Model-Based Investigation of the Biological Origin of Human Social Perception of Faces. CogSci 2019. [pdf]
Sophia J. Huang, Chaitanya K. Ryali, Jianling Liu, Dalin Guo, Jinyan Guan, Yvonne Li, Angela J. Yu. (2019)

Why so gloomy? A Bayesian explanation of human pessimism bias in the multi-armed bandit task. NeurIPS 2018. [pdf] [poster]
Dalin Guo, Angela J. Yu. (2018)

Anhedonia and anxiety underlying depressive symptomatology have distinct effects on reward-based decision-making. PLoS ONE, 12(10):e0186473. [pdf]
Katia M. Harle, Dalin Guo, Shunan Zhang, Martin P. Paulus, Angela J. Yu. (2017)

Conference Presentation

Contributions of reward and uncertainty to human choice behavior in a novel sequential decision-making task.
the BRAIN Initiative Investigators Meeting, virtual.
Dalin Guo, Alexander M. Paunov, Florent Meyniel, Angela J. Yu (2021)

Uncertainty-driven human exploration in a novel foraging task.
the BRAIN Initiative Investigators Meeting, virtual.
Dalin Guo, Florent Meyniel, Angela J. Yu (2020)

Features or Bugs? Synergistic Idiosyncrasies in Human Learning and Decision-Making.
Unifying AI and Neuroscience Québec Centre Student Symposium. (lightning talk) [video (Youtube)]

Features or Bugs? Synergistic Idiosyncrasies in Human Learning and Decision-Making.
Neurochat online conference. (Short talk in Chinese) [video (Bilibili)] [video (Youtube)]

Features or Bugs? Synergistic Idiosyncrasies in Human Learning and Decision-Making.
Neuromatch 1.0 online conference. (Short talk)

Evidence for uncertainty-driven human exploration in a novel multi-armed bandit task.
CoSyNe Abstracts 2020, Denver, CO, USA.
Angela J. Yu, Dalin Guo, Florent Meyniel (2020)

Deep Bayesian Bandits for Online Personalized Recommendations.
14th Women in Machine Learning Workshop, Vancouver, CA.
Dalin Guo, Sofia Ira Ktena, Ferenc Huszar, Wenzhe Shi, Alykhan Tejani. (2019)

Human learning and decision-making in the bandit task: Three wrongs make a right
Conference on Cognitive Computational Neuroscience, Berlin, Germany
Dalin Guo, Angela J. Yu. (2019)

Why so gloomy? A Bayesian explanation of human pessimism bias in the multi-armed bandit task.
13th Women in Machine Learning Workshop, Montreal, CA.
Dalin Guo, Angela J. Yu. (2018)

Humans underestimate reward probability and overestimate environmental volatility in a multi-armed bandit task - insights from a Bayesian analysis of human learning and decision-making.
SfN Annual Meeting, San Diego, CA, USA. (Short talk)
Dalin Guo, Angela J. Yu. (2018)

Recovering human reward expectation in a bandit setting using Bayesian models.
CoSyNe Abstracts 2018, Denver, CO, USA.
Dalin Guo, Florent Meyniel, Angela J. Yu. (2018)

Human learning and decision-making in the multi-armed bandit task.
12th Women in Machine Learning Workshop, Long Beach, CA, USA .
Dalin Guo, Angela J. Yu. (2017)

Dependence of reward-based learning and decision-making on environmental statistics such as reward abundance and variance.
SfN Annual Meeting, Washington D.C., USA.
Dalin Guo, Angela J. Yu. (2017)

Human learning and decision-making in the multi-armed bandit task.
SoCal Machine Learning Symposium, Los Angeles, CA, USA.
Dalin Guo, Angela J. Yu. (2017)

Services

Reviewer: Neurips 2019, 2020, 2021; ICLR 2020, 2021; WiML 2017, 2018, 2019; CCN 2019;
Area Chair: WiML 2019, 2020, 2021
Volunteer: Neurips 2017, 2019; Bernstein Conference 2019; Neuromatch Academy 2020

Teaching

COGS 3. Intro to Computing - Fall 2019
COGS 178. Genes, Brains, and Behavior - Spring 2019, 2021
COGS 118A. Supervised Machine Learning Algorithms - Fall 2018
COGS 118D. Mathematical Statistics - Winter 2018, 2017

Resources

Here are some notes/slides I made during grad school:
(Online) Meta-learning
Variational Autoencoder (VAE)
Sparse Coding
Robust PCA