Sebastian Farquhar - Unbiased Active Learning and Testing

Date:

September 16, 2022

Author:

Hrvoje Stojic

Unbiased Active Learning and Testing



Abstract

Active learning is a powerful tool when labelling data is expensive, but it introduces a bias because the training data no longer follows the population distribution. We can, in fact, remove this bias using corrective weights based on importance sampling. This has two main consequences: first, we show that the bias is actually useful for active learning, especially with overparameterized models like neural networks; second, this technique enables active testing---a new way of doing model evaluation with limited data.


Notes


Share on social media

Share on social media

Share on social media

Share on social media

Related Seminars

Mickael Binois - Leveraging replication in active learning

We were recently joined by Mickael Binois, to talk about 'Leveraging replication in active learning'.

Jun 24, 2024

Mickael Binois - Leveraging replication in active learning

We were recently joined by Mickael Binois, to talk about 'Leveraging replication in active learning'.

Jun 24, 2024

Mickael Binois - Leveraging replication in active learning

We were recently joined by Mickael Binois, to talk about 'Leveraging replication in active learning'.

Jun 24, 2024

Mickael Binois - Leveraging replication in active learning

We were recently joined by Mickael Binois, to talk about 'Leveraging replication in active learning'.

Jun 24, 2024

Ilija Bogunovic - From Data to Confident Decisions

We were recently joined by Ilija Bogunovic, to talk about 'Robust and Efficient Algorithmic Decision Making'.

Jun 13, 2024

Ilija Bogunovic - From Data to Confident Decisions

We were recently joined by Ilija Bogunovic, to talk about 'Robust and Efficient Algorithmic Decision Making'.

Jun 13, 2024

Ilija Bogunovic - From Data to Confident Decisions

We were recently joined by Ilija Bogunovic, to talk about 'Robust and Efficient Algorithmic Decision Making'.

Jun 13, 2024

Ilija Bogunovic - From Data to Confident Decisions

We were recently joined by Ilija Bogunovic, to talk about 'Robust and Efficient Algorithmic Decision Making'.

Jun 13, 2024

Dario Azzimonti - Preference learning with Gaussian processes

We were recently joined by Dario Azzimonti, to talk about 'Preference learning with Gaussian processes'.

May 23, 2024

Dario Azzimonti - Preference learning with Gaussian processes

We were recently joined by Dario Azzimonti, to talk about 'Preference learning with Gaussian processes'.

May 23, 2024

Dario Azzimonti - Preference learning with Gaussian processes

We were recently joined by Dario Azzimonti, to talk about 'Preference learning with Gaussian processes'.

May 23, 2024

Dario Azzimonti - Preference learning with Gaussian processes

We were recently joined by Dario Azzimonti, to talk about 'Preference learning with Gaussian processes'.

May 23, 2024

Mojmír Mutný - Optimal Experiment Design in Markov Chains

We were recently joined by Mojmír Mutný (ETH Zurich), to talk about 'Optimal Experiment Design in Markov Chains'.

Mar 28, 2024

Mojmír Mutný - Optimal Experiment Design in Markov Chains

We were recently joined by Mojmír Mutný (ETH Zurich), to talk about 'Optimal Experiment Design in Markov Chains'.

Mar 28, 2024

Mojmír Mutný - Optimal Experiment Design in Markov Chains

We were recently joined by Mojmír Mutný (ETH Zurich), to talk about 'Optimal Experiment Design in Markov Chains'.

Mar 28, 2024

Mojmír Mutný - Optimal Experiment Design in Markov Chains

We were recently joined by Mojmír Mutný (ETH Zurich), to talk about 'Optimal Experiment Design in Markov Chains'.

Mar 28, 2024