ConceptAI & ML
Overfitting
Overfitting occurs when a model fits the training data too well, capturing noise instead of the underlying pattern, leading to poor generalization on out-of-sample data.
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
9Each in their own terms, from the easiest level.
Pieter Abbeel
Overfitting occurs when you stop modeling the patterns in the training data that generalize and start modeling the noise that doesn't.Yaser Abu-Mostafa
Fitting the data more than is warranted, at the expense of generalisation out of sample.Raj Chetty
Designing a model that works well within a specific training dataset but fails to predict well on new, unseen data.Andrej Karpathy
A modeling error that occurs when a function is too closely fit to a limited set of data points, resulting in low training loss.Phillip Isola
The model has learned noise in the training set rather than the underlying pattern, leading to poor generalization on unseen data. \n- Prerequisites: Overfitting, generalization, train/test errorBrian Yu
A phenomenon where a model fits too closely to a particular training data set and therefore fails to generalize to future data.Gilbert Strang
A modeling error that occurs when a function is too closely fit to a limited set of data points.Shahar Mendelson
A modeling error that occurs when a function is too rich and fits the training data too closely at the expense of generalization.Israel Borokini
A modeling state where a model fits the training data too closely, capturing noise and failing to generalize well to test data.
Lectures that assume it
17The first 12 are shown.
- Let's Build GPT: From Scratch, in Code, Spelled Out
- Training Neural Networks II
- Validation
- CNN Architectures
- Three Learning Principles
- Mechanism Design & Game-Theoretic Incentives for Collaborative Learning
- Transfer Learning: Models
- ImageNet and Convolutional Neural Networks
- Loss Functions and Optimization
- Efficient Methods and Hardware for Deep Learning
- Hacker's Guide to Deep Learning
- Generative Models: Basics
Connections
40The lectures that define it or assume it, and the concepts next to it.
Concepts9
Show 3 moreShow fewer
Lectures31
- Building makemore Part 2: Multi-Layer Perceptron Language Models (Bengio et al. 2003)
- Harvard CS50’s Artificial Intelligence with Python – Full University Course
- Overfitting
- Regularization
- The Economics of Health Care and Insurance
- Improving Judicial Decisions
Show 25 moreShow fewer
- Machine Learning: Naive Bayes
- Machine Learning: Perceptrons and Logistic Regression
- Saddle Points Continued, Maxmin Principle
- AI in Science and Engineering: The Role of Academia in Understanding AI
- Generalization Theory
- Machine Learning: Neural Networks and Decision Trees
- Reinforcement Learning Part 2/2
- CDFW Conservation Lecture Series: November 2020
- Let's Build GPT: From Scratch, in Code, Spelled Out
- Training Neural Networks II
- Validation
- CNN Architectures
- Three Learning Principles
- Mechanism Design & Game-Theoretic Incentives for Collaborative Learning
- Transfer Learning: Models
- ImageNet and Convolutional Neural Networks
- Loss Functions and Optimization
- Efficient Methods and Hardware for Deep Learning
- Hacker's Guide to Deep Learning
- Generative Models: Basics
- Transformers in Vision: Tackling problems in Computer Vision
- Introduction to Deep Learning
- Backpropagation: Finding Partial Derivatives
- Emergent Behaviors in Foundational Models
- Current & Future Application of AI in Epilepsy by Dr. Barbara Kwiecinska, PGY4 Neurology Resident