ConceptAI & ML
Generalization
The ability of a learning model to perform accurately on unseen data outside the training set.
Know first
Named as a prerequisite by the lectures that define this. The number is how many say so. A grey one has no page yet.
Opens up
Concepts that name this one as a prerequisite.
Lecturers
7Each in their own terms, from the easiest level.
Yaser Abu-Mostafa
Emma Brunskill
The ability to map past experiences to actions in extremely large or unseen scenarios.Patrick Winston
Picking some pieces in the phoneme matrix that you just don't care about.Phillip Isola
How well a trained neural network performs on new, unseen data from the world.Pieter Abbeel
The process of applying knowledge learned from a limited set of training states to new, unseen, but related states.Stefano Ermon
The ability of a generative model to produce samples that are not present in the training data but are consistent with the learned distribution.François Chollet
The ability to perform well on tasks or situations that the system has not been explicitly trained on.
Lectures that assume it
16The first 12 are shown.
- The VC Dimension
- Adversarial Examples and Adversarial Training
- DeepMind's Perceiver and Perceiver IO: New data family architecture
- Bias-Variance Tradeoff
- Neural Networks
- Three Learning Principles
- From System 1 to System 2 Deep Learning
- Benchmarking and Evaluation: How Do We Know What LLMs Can Do?
- Machine Learning: Naive Bayes
- AI in Science and Engineering: The Role of Academia in Understanding AI
- Machine Learning: Neural Networks and Decision Trees
- Challenges and Open Problems in Deep Reinforcement Learning (CS 285, Lecture 23, Part 1)
Connections
34The lectures that define it or assume it, and the concepts next to it.
Concepts10
Show 4 moreShow fewer
Lectures24
- Generative Modeling by Estimating Gradients of the Data Distribution
- The Learning Problem
- Regularization
- On the Measure of Intelligence: Program Synthesis and Fluid Abstraction
- Introduction to Reinforcement Learning (CS234, 2024, Lecture 1)
- Learning: Sparse Spaces and Phonological Generalization
Show 18 moreShow fewer
- Generalization Theory
- Reinforcement Learning Part 2/2
- The VC Dimension
- Adversarial Examples and Adversarial Training
- DeepMind's Perceiver and Perceiver IO: New data family architecture
- Bias-Variance Tradeoff
- Neural Networks
- Three Learning Principles
- From System 1 to System 2 Deep Learning
- Benchmarking and Evaluation: How Do We Know What LLMs Can Do?
- Machine Learning: Naive Bayes
- AI in Science and Engineering: The Role of Academia in Understanding AI
- Machine Learning: Neural Networks and Decision Trees
- Challenges and Open Problems in Deep Reinforcement Learning (CS 285, Lecture 23, Part 1)
- Questions for Theory in the New Age of Machine Learning
- Conceptual Abstraction and Analogy in Natural and Artificial Intelligence
- Introduction to Deep Learning
- Emergent Behaviors in Foundational Models