Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

Reinforcement01:23

Reinforcement

308
Positive and negative reinforcement are key concepts in operant conditioning, a learning process where the consequences of a behavior affect the likelihood of that behavior being repeated.
Positive reinforcement occurs when a behavior is followed by the presentation of a rewarding stimulus, increasing the frequency of that behavior. For example:
308
Reinforcement Schedules01:24

Reinforcement Schedules

227
Positive reinforcement is a powerful method for teaching new behaviors to both animals and humans. B.F. Skinner demonstrated this with his experiments using rats in a Skinner box. When a rat pressed a lever, it received a food pellet. This immediate reward encouraged the rat to repeat the behavior. This method, where a reward follows every instance of the behavior, is known as continuous reinforcement. It is highly effective for establishing new behaviors quickly.
Once a behavior is learned,...
227
Observational Learning01:12

Observational Learning

259
Albert Bandura's observational learning, also known as imitation or modeling, occurs when a person observes and imitates another's behavior. It is a quicker process than operant conditioning. A well-known example is the Bobo doll study, where children who saw an adult acting aggressively towards the doll were more likely to act aggressively when left alone, compared to those who observed a nonaggressive adult. Many psychologists view observational learning as a form of latent learning...
259
Associative Learning01:27

Associative Learning

503
Associative learning is a fundamental concept in behavioral psychology, wherein a connection is established between two stimuli or events, leading to a learned response. This process is critical in understanding how behaviors are acquired and modified. Conditioning, the mechanism through which associations are formed, can be divided into two main types: classical conditioning and operant conditioning, each elucidating different aspects of associative learning.
Classical conditioning, also known...
503
State Space Representation01:27

State Space Representation

260
The frequency-domain technique, commonly used in analyzing and designing feedback control systems, is effective for linear, time-invariant systems. However, it falls short when dealing with nonlinear, time-varying, and multiple-input multiple-output systems. The time-domain or state-space approach addresses these limitations by utilizing state variables to construct simultaneous, first-order differential equations, known as state equations, for an nth-order system.
Consider an RLC circuit, a...
260
Purposive Learning01:22

Purposive Learning

181
E. C. Tolman emphasized the purposiveness of behavior — the idea that much of our behavior is goal-directed. For instance, employees who aim for a promotion work diligently to meet their targets. Tolman argued that when classical conditioning and operant conditioning occur, the organism acquires certain expectations. In classical conditioning, a child might fear a dog because they expect it to bite. In operant conditioning, a person might consistently work overtime because they expect a...
181

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Broad-spectrum empirical antibiotic overuse in community-onset bacteremia: prevalence, outcomes, and associated factors.

International journal of antimicrobial agents·2026
Same author

Valorizing ginseng residues using dimethyl ether: Recovering functional lipids and ginsenosides.

Journal of ginseng research·2026
Same author

Oncological Outcomes of Nonmuscle-Invasive Bladder Cancer in Patients Aged 45 Years and Younger: A Retrospective Matched Cohort Study.

The Journal of urology·2026
Same author

Identification of the optimal manipulation medium temperature for in vitro handling of oocytes and embryos during in vitro maturation, parthenogenesis, and in vitro culture in pigs.

The Journal of reproduction and development·2026
Same author

Household-level surrounding greenspace as a nature-based intervention for health recovery after occupational injury.

Frontiers in public health·2026
Same author

Axial-time mapping: A diagnostic method to reveal concealed long-term catalyst deactivation mechanism in CO<sub>2</sub> hydrogenation.

Science advances·2026

Related Experiment Video

Updated: Aug 15, 2025

An Open-Source Virtual Reality System for the Measurement of Spatial Learning in Head-Restrained Mice
08:59

An Open-Source Virtual Reality System for the Measurement of Spatial Learning in Head-Restrained Mice

Published on: March 3, 2023

2.2K

STACoRe: Spatio-temporal and action-based contrastive representations for reinforcement learning in Atari.

Young Jae Lee1, Jaehoon Kim1, Mingu Kwak2

  • 1School of Industrial and Management Engineering, Korea University, Seoul, Republic of Korea.

Neural Networks : the Official Journal of the International Neural Network Society
|January 1, 2023
PubMed
Summary

This study introduces spatio-temporal and action-based contrastive representation (STACoRe) learning to improve sample efficiency in deep reinforcement learning (DRL). STACoRe enhances agent learning with limited interactions by effectively capturing semantic representations from image-based environments.

Keywords:
AtariAutomatic data augmentationEnd-to-end learningReinforcement learningSpatio-temporal contrastive learningSupervised contrastive learning

More Related Videos

WheelCon: A Wheel Control-Based Gaming Platform for Studying Human Sensorimotor Control
08:18

WheelCon: A Wheel Control-Based Gaming Platform for Studying Human Sensorimotor Control

Published on: August 15, 2020

5.0K
Combining Computer Game-Based Behavioural Experiments With High-Density EEG and Infrared Gaze Tracking
13:40

Combining Computer Game-Based Behavioural Experiments With High-Density EEG and Infrared Gaze Tracking

Published on: December 16, 2010

16.8K

Related Experiment Videos

Last Updated: Aug 15, 2025

An Open-Source Virtual Reality System for the Measurement of Spatial Learning in Head-Restrained Mice
08:59

An Open-Source Virtual Reality System for the Measurement of Spatial Learning in Head-Restrained Mice

Published on: March 3, 2023

2.2K
WheelCon: A Wheel Control-Based Gaming Platform for Studying Human Sensorimotor Control
08:18

WheelCon: A Wheel Control-Based Gaming Platform for Studying Human Sensorimotor Control

Published on: August 15, 2020

5.0K
Combining Computer Game-Based Behavioural Experiments With High-Density EEG and Infrared Gaze Tracking
13:40

Combining Computer Game-Based Behavioural Experiments With High-Density EEG and Infrared Gaze Tracking

Published on: December 16, 2010

16.8K

Area of Science:

  • Artificial Intelligence
  • Machine Learning
  • Deep Reinforcement Learning

Background:

  • Deep reinforcement learning (DRL) agents learn through interaction but often require excessive data, leading to sample inefficiency.
  • Existing methods for sample-efficient DRL primarily focus on visual similarity, struggling to capture essential semantic representations.
  • Effective state representation learning is crucial for improving DRL performance with limited environmental interactions.

Purpose of the Study:

  • To propose a novel method, spatio-temporal and action-based contrastive representation (STACoRe) learning, to enhance sample efficiency in DRL.
  • To develop a DRL approach that effectively learns semantic state representations from limited interactions.
  • To improve the performance of DRL agents in image-based environments with restricted data.

Main Methods:

  • STACoRe employs dual contrastive learning strategies: one utilizing agent actions as pseudo-labels and the other leveraging spatio-temporal information.
  • Action-based contrastive learning incorporates an automated selection of environment-specific data augmentation techniques for stable training.
  • The model is trained end-to-end by simultaneously optimizing action-based and spatio-temporal contrastive loss functions.

Main Results:

  • STACoRe demonstrates superior sample efficiency compared to existing methods in deep reinforcement learning.
  • Experiments conducted on 26 Atari 2600 benchmark games with limited interaction (100k steps) validate the proposed approach.
  • The method effectively learns improved state representations, leading to better performance in sample-constrained scenarios.

Conclusions:

  • The proposed STACoRe learning method significantly enhances sample efficiency in deep reinforcement learning.
  • By integrating action-based and spatio-temporal contrastive learning, STACoRe effectively captures semantic representations crucial for DRL.
  • This approach offers a promising direction for developing more data-efficient intelligent agents in complex environments.