Related Experiment Videos
Trigger-based conditional selective unlearning with elastic weight consolidation
Hyun Kwon1, Joo Bon Maeng1, Dae-Jin Kim2
1Department of Artificial Intelligence and Data Science, Korea Military Academy, Seoul, 01805, South Korea.
Scientific Reports
|July 18, 2026
Summary
This study introduces a novel trigger-based method for machine learning unlearning, enabling data deletion without retraining. The approach effectively removes specific training data influence, crucial for privacy compliance and data governance.
Area of Science:
- Machine Learning
- Data Privacy
- Artificial Intelligence
Background:
- Data deletion requests pose challenges for trained machine learning models.
- Retraining models from scratch is computationally expensive and impractical.
- Efficient unlearning methods are essential for privacy compliance and responsible AI.
Purpose of the Study:
- To develop a method for removing specific training samples' influence from a trained classifier on demand.
- To implement a trigger-based unlearning framework that avoids retraining.
- To evaluate the effectiveness and selectivity of the proposed unlearning technique.
Main Methods:
- A three-stage training pipeline: Pretraining, Conditioning (binding trigger to uniform-posterior response), and Unlearning (applying uniform target with Elastic Weight Consolidation).
- Utilized a trigger-based framing where input patterns activate forgetting behavior at inference time.
- Evaluated on CIFAR-10 and SVHN datasets using a comprehensive six-condition protocol.
Main Results:
- The proposed model achieved significant conditional gaps (23.04 pp on CIFAR-10, 71.54 pp on SVHN), demonstrating effective input-conditioned unlearning.
- Membership inference attack AUC scores were close to chance (0.502 on CIFAR-10, 0.511 on SVHN), indicating successful data removal.
- A residual clean-input capacity loss of ~17 pp (CIFAR-10) and ~9 pp (SVHN) was observed as a limitation.
Conclusions:
- The trigger-based unlearning method effectively removes specific training data influence without retraining.
- The framework offers a selective and efficient approach to data deletion in machine learning models.
- Further research is needed to address the residual clean-input capacity loss for complete model integrity.
Related Concept Videos
01:27Associative Learning
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...
Classical conditioning, also known...
01:14Avoidance Learning and Learned Helplessness
Avoidance learning and learned helplessness are critical concepts in understanding behavioral responses to negative stimuli.
Avoidance learning occurs when an organism learns that a specific behavior can prevent an unpleasant outcome. For example, a student who receives a bad grade may start studying harder to avoid future poor grades. This behavior persists even when the negative outcome is no longer present. Avoidance learning is powerful because it maintains behavior in the absence of the...
Avoidance learning occurs when an organism learns that a specific behavior can prevent an unpleasant outcome. For example, a student who receives a bad grade may start studying harder to avoid future poor grades. This behavior persists even when the negative outcome is no longer present. Avoidance learning is powerful because it maintains behavior in the absence of the...
01:12Observational Learning
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 because...
01:21Cognitive Learning
Cognitive learning is based on purposive behavior, incidental learning, and insight learning.
E. C. Tolman's theory of purposive behavior emphasizes that much behavior is goal-directed. He argued that to understand behavior, we must look at the entire sequence of actions leading to a goal. For instance, high school students study hard, not just due to past reinforcement but also to achieve the goal of getting into a good college.
Tolman introduced the idea that behavior is influenced by...
E. C. Tolman's theory of purposive behavior emphasizes that much behavior is goal-directed. He argued that to understand behavior, we must look at the entire sequence of actions leading to a goal. For instance, high school students study hard, not just due to past reinforcement but also to achieve the goal of getting into a good college.
Tolman introduced the idea that behavior is influenced by...
01:24Generalization, Discrimination, and Extinction
Generalization, discrimination, and extinction are key concepts in operant conditioning that influence how behaviors are learned and maintained.
Generalization occurs when a behavior reinforced in one context is performed in similar situations. For instance, a student who studies diligently for calculus and receives excellent grades might apply the same study habits to psychology and history, expecting similar results. Generalization shows how learning in one setting can influence behavior in...
Generalization occurs when a behavior reinforced in one context is performed in similar situations. For instance, a student who studies diligently for calculus and receives excellent grades might apply the same study habits to psychology and history, expecting similar results. Generalization shows how learning in one setting can influence behavior in...
01:22Purposive Learning
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 bonus...