Curriculum learning
Explore this paper's citation graph
Summary
It is hypothesized that curriculum learning has both an effect on the speed of convergence of the training process to a minimum and on the quality of the local minima obtained: curriculum learning can be seen as a particular form of continuation method (a general strategy for global optimization of non-convex functions).
- Type
- article
- Published
- 2009-06-14
- Cited by
- 7,538
- References
- 33
- Access
- Open access
- OpenAlex
- https://openalex.org/W2296073425
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:873046
Keywords
Curriculum, Computer science, Generalization, Context (archaeology), Maxima and minima
References
- Learning a Nonlinear Embedding by Preserving Class Neighbourhood Structure
- Explanation-based neural network learning a lifelong learning approach
- Language acquisition in the absence of explicit negative evidence: how important is starting small?
- Generalization in the programed teaching of a perceptron.
- A day of great illumination: B. F. Skinner's discovery of shaping.
- An empirical evaluation of deep architectures on problems with many factors of variation
- Flexible shaping: how learning in small steps helps.
- Extracting and composing robust features with denoising autoencoders
- Connectionist language modeling for large vocabulary continuous speech recognition
- Parallel continuation-based global optimization for molecular conformation and protein folding
- Learning Deep Architectures for AI
- Numerical continuation methods - an introduction
- Global Continuation for Distance Geometry Problems
- Restricted Boltzmann machines for collaborative filtering
- On the power of small-depth threshold circuits
- Sparse Feature Learning for Deep Belief Networks
- A unified architecture for natural language processing: deep neural networks with multitask learning
- The Difficulty of Training Deep Architectures and the Effect of Unsupervised Pre-Training
- A Fast Learning Algorithm for Deep Belief Nets
- Neural network learning control of robot manipulators using gradually increasing task difficulty
Cited by
- Language Models With Meta-information
- The developing infant creates a curriculum for statistical learning
- Towards large-scale multimedia retrieval enriched by knowledge about human interpretation
- The Role of Sequences for Incremental Learning
- Compositional Matrix-Space Models for Sentiment Analysis
- Three Dependency-and-Boundary Models for Grammar Induction
- Learning to see like children: proof of concept
- New Algorithms for Large-Scale Support Vector Machines. (Nouveaux Algorithmes pour l'Apprentissage de Machines à Vecteurs Supports sur de Grandes Masses de Données)
- Scalable Lifelong Learning with Active Task Selection
- Learning in modular systems
- Efficiently Learning Nonlinear Classifiers for Domain Specific Performance Measures
- I2VM: Incremental import vector machines
- Learning domain abstractions for long lived robots
- Active Task Selection for Lifelong Machine Learning
- Incremental on-line adaptation of POMDP-based dialogue managers to extended domains
- Never-Ending Learning
- A fast and efficient pre-training method based on layer-by-layer maximum discrimination for deep neural networks
- Apprentissage de représentations et robotique développementale : quelques apports de l'apprentissage profond pour la robotique autonome. (Representation learning and developmental robotics : on the use of deep learning for autonomous robots)
- Learning Deep Representations : Toward a better new understanding of the deep learning paradigm. (Apprentissage de représentations profondes : vers une meilleure compréhension du paradigme d'apprentissage profond)
- Learning generative models of mid-level structure in natural images
Related papers
- DETERMINING QUALITY REQUIREMENTS AT THE UNIVERSITIES TO IMPROVE THE QUALITY OF EDUCATION
- Faster escaping from local minima for back-propagation algorithm
- A USEFUL BP ALGORITHM TO OVERCOME LOCAL MINIMA
- Eliminating all bad Local Minima from Loss Landscapes without even adding an Extra Unit
- Locating the local minima in lens design with machine learning
- Building and navigating a network of local minima
- Avoiding false local minima by proper initialization of connections