How Do Humans Teach: On Curriculum Learning and Teaching Dimension

By Faisal Khan, Xiaojin Zhu, Bilge Mutlu. 2011

188 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Machine Teaching: An Inverse Problem to Machine Learning and an Approach Toward Optimal Education (2015, methodology): “This is the strategy used in (Khan, Zhu, and Mutlu 2011).”
Knowledge Matters: Importance of Prior Information for Optimization (2013, background): “Recent experiments on human subjects also indicates that humans teach by using a curriculum strategy (Khan et al., 2011).”
Machine Teaching for Bayesian Learners in the Exponential Family (2013, methodology): “In addition, our optimal teaching framework may shed light on the optimality of different method of teaching humans [9, 13, 17, 18].”
Learning something from nothing: Leveraging implicit human feedback strategies (2014, background): “There is also a growing body of work that examines how humans want to teach agents by providing demonstrations in a sequential decision task [4], or by selecting a sequence of data in a classification task [7].”
Learning behaviors via human-delivered discrete feedback: modeling implicit feedback strategies to speed up learning (2015, background): “Within the field of artificial intelligence there exists a significant body of work on the problem of designing agents to learn behaviors from human trainers [5,14,27], and specifically on the problem of learning from trainer-provided feedback [8,17].”
Algorithmic and Human Teaching of Sequential Decision Tasks (2012, background): “(Khan, Zhu, and Mutlu 2011) demonstrated that this is the most prominent strategy in human teaching and provided a theoretical account for it.”
Self-Paced Curriculum Learning (2015, methodology): “From human behavioral perspective , evidence have shown that CL is consistent with the principle in human teaching (Khan, Zhu, and Mutlu 2011; Basu and Christensen 2013).”
Optimizing Instructional Policies (2013, background): “Theorists have also made computational arguments for the benefit of fading (Bengio, Louradour, Collobert, & Weston, 2009; Khan et al., 2011).”
Teacher Improves Learning by Selecting a Training Subset (2018, background): “Machine teaching applications include education [14, 33, 39, 27, 13, 34], computer security [2, 1, 32], and interactive machine learning [40, 12, 24].”
Breaking the Curse of Space Explosion: Towards Efficient NAS with Curriculum Search (2020, background): “Khan et al. (2011) provide evidence about the consistency between curriculum learning and the learning principles of human.”
Easy Questions First? A Case Study on Curriculum Learning for Question Answering (2016, background): “Curriculum learning is also related to teaching dimension (Khan et al., 2011) which studies the strategies that humans follow as they teach a target concept to a robot by assuming a teaching goal of minimizing the learner’s expected generalization error at each iteration.”
Neural Networks and Deep Learning (2018, unknown): “The basic principles of curriculum learning are discussed in [238].”
Neural Networks for Sequential Data: a Pre-training Approach based on Hidden Markov Models (2015, background): “The rationale behind this proposal is that human learning is facilitated if we start to approach simpler concepts before dealing with more complex ones [44].”
Curriculum Learning for Reinforcement Learning Domains: A Framework and Survey (2020, background): “To study the teaching strategies followed by naive users, Khan et al. (2011) conduct behavioral studies in which human participants need to teach a robot the concept of whether an object can be grasped with one hand.”
Curriculum Learning by Transfer Learning: Theory and Experiments with Deep Networks (2018, background): “We do not delve any deeper into the question of machine teaching, but see (Goldman & Kearns, 1995; Khan et al., 2011).”
Self-Paced Learning with Diversity (2014, background): “[10] provided evidence that CL is consistent with the principle in teaching.”