Machine Learning Blog | ML@CMU | Carnegie Mellon University

machine learning Research

Document Grounded Generation

by Shrimai Prabhumoye / June 25, 2021

Figure 1: Document Grounded Generation – An example of a conversation that is grounded in the given document (text in green shows information from the document that was used to generate the response). Natural language generation (NLG) systems are increasingly expected to be naturalistic, content-rich, and situation-aware due to their popularity and pervasiveness in human life. This is particularly relevant in dialogue systems, machine translation systems, story generation, and question answering systems. Despite these mainstream applications, NLG systems face the…

19 1265

machine learning Research

See, Hear, Explore: Curiosity via Audio-Visual Association

June 11, 2021

See, Hear, Explore: We propose a formulation of curiosity that encourages the agent to explore novel associations between modalities, such as audio and vision. In Habitat, the photorealistic navigation simulator shown above, our method allows for more efficient exploration than…

31 1915

machine learning Research

Decentralized Multi-Robot Active Search

June 4, 2021

Active Search (Active Sensing) in robotics refers to the problem of searching for objects of interest in an unknown environment by making data-collection decisions. Active Search has many applications including the search and rescue of human survivors following disasters, detecting…

42 3249

artificial intelligence machine learning reinforcement learning Research

PLAS: Latent Action Space for Offline Reinforcement Learning

April 30, 2021

Figure 1: Overview: To avoid out-of-distribution actions in the Offline Reinforcement Learning problem, we propose to implicitly constrain the policy by modifying the action space instead of enforcing explicit constraints. The policy is trained to output a latent action which…

62 1356

machine learning Research

A Learning Theoretic Perspective on Local Explainability

April 23, 2021

Fig 1: A formal relationship between interpretability and complexity. Going from left to right, we consider increasingly complex functions. As the complexity increases, local linear explanations can approximate the function only in smaller and smaller neighborhoods. These neighborhoods, in other…

70 4938

machine learning Research

Counterfactual predictions under runtime confounding

April 16, 2021

Due to feasibility or ethical requirements, a prediction model may only access a subset of the confounding factors that affect both the decision and outcome. We propose a procedure for learning valid counterfactual predictions in this setting.…

78 3361

FATE machine learning Research

Tilted Empirical Risk Minimization

April 2, 2021

In this post, we describe a simple extension to empirical risk minimization---tilted ERM (TERM), which can be used for enforcing fairness between subgroups, mitigating the effect of outliers, and handling class imbalance, all in a unified framework.…

67 3021

machine learning Research

Carnegie Mellon University at the Conference on Fairness, Accountability, and Transparency (ACM FAccT 2021)

March 5, 2021

This week researchers from all across computer science, social sciences and humanities are gathering for the flagship conference of the emerging field of Fairness, Accountability and Transparency in algorithmic systems: FAccT. FAccT (previously FAT*) is dedicated to the inherent risks…

88 1948

deep learning federated learning machine learning Research

An Inferential Perspective on Federated Learning

February 19, 2021

TL;DR: motivated to better understand the fundamental tradeoffs in federated learning, we present a probabilistic perspective that generalizes and improves upon federated optimization and enables a new class of efficient federated learning algorithms.…

80 5987

machine learning Research

Representational Aspects of Depth and Conditioning in Normalizing Flows

December 18, 2020

Top and Bottom Right: RealNVP [3] uses checkerboard and channel-wise partitioning schemes in order to factor out parameters and ensure that there aren’t redundant partitions from previous layers. GLOW [4] uses an invertible 1×1 convolution which allows the partitioned to…

95 5071

Older Posts

Machine Learning Blog | ML@CMU | Carnegie Mellon University

Statistics:

Categories: