Detecting Individual Decision-Making Style: Exploring Behavioral Stylometry in Chess
Dec 6, 2021
Speakers
Reid McIlroy-Young
Speaker · 0 followers
Russell Wang
Speaker · 0 followers
Siddhartha Sen
Speaker · 0 followers
Jon Kleinberg
Speaker · 0 followers
Ashton Anderson
Speaker · 0 followers
About
The advent of machine learning models that surpass human decision-making ability in complex domains has initiated a movement towards building AI systems that can interact with humans. Many building blocks are essential to this activity, and one of the central ones is the algorithmic characterization of aspects of human behavior. While much of the existing work focuses on characterizing human behavior in an aggregate or generic sense, an important longer-range goal is to specialize these characterizations to individual people. Specializing behavioral models to individual people requires that we be able to characterize what differentiates specific people. As a means of formalizing this process, we study the problem of behavioral stylometry, in which the task is to identify a decision-maker from their decisions alone. We present a transformer-based approach to behavioral stylometry in the context of chess, in which a game's players are identified from their moves. Our method operates in a few-shot classification framework, and can correctly identify a player from among thousands of candidate players given only 100 labeled games with up to 98The advent of machine learning models that surpass human decision-making ability in complex domains has initiated a movement towards building AI systems that can interact with humans. Many building blocks are essential to this activity, and one of the central ones is the algorithmic characterization of aspects of human behavior. While much of the existing work focuses on characterizing human behavior in an aggregate or generic sense, an important longer-range goal is to specialize these characte…
Organizer
NeurIPS 2021
Account · 1.9k followers
About NeurIPS 2021
Neural Information Processing Systems (NeurIPS) is a multi-track machine learning and computational neuroscience conference that includes invited talks, demonstrations, symposia and oral and poster presentations of refereed papers. Following the conference, there are workshops which provide a less formal setting.
Like the format? Trust SlidesLive to capture your next event!
Professional recording and live streaming, delivered globally.
Sharing
Recommended Videos
Presentations on similar topic, category or speaker
Partition-Based Formulations for Mixed-Integer Optimization of Trained ReLU Neural Networks
Watch later
Calvin Tsay, …
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
Machine Learning for Variance Reduction in Online Experiments
Watch later
Yongyi Guo, …
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
Improved Object Detection in Thermal Imaging Through Context Enhancement and Information Fusion: A Case Study in Autonomous Driving
Watch later
Junchi Bin, …
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
Not All Low-Pass Filters are Robust in Graph Convolutional Networks
Watch later
Heng Chang, …
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
Live Q&A Session 2
Watch later
Susan Athey, …
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
Simulated User Studies for Explanation Evaluation
Watch later
Valerie Chen, …
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%