An official website of the United States government
The performance of artificial neural networks (ANNs) degrades when training data are limited or imbalanced. In contrast, the human brain can learn quickly from just a few examples. Here, we investigated the role of sleep in improving the performance of ANNs trained with limited data on the MNIST and Fashion MNIST datasets. Sleep was implemented as an unsupervised phase with local Hebbian type learning rules. We found a significant boost in accuracy after the sleep phase for models trained with limited data in the range of 0.5-10% of total MNIST or Fashion MNIST datasets. When more than 10% of the total data was used, sleep alone had a slight negative impact on performance, but this was remedied by fine-tuning on the original data. This study sheds light on a potential synaptic weight dynamics strategy employed by the brain during sleep to enhance memory performance when training data are limited or imbalanced.
more »
« less
Unsupervised Replay Strategies for Continual Learning with Limited Data
Artificial neural networks (ANNs) show limited performance with scarce or imbalanced training data and face challenges with continuous learning, such as forgetting previously learned data after new tasks training. In contrast, the human brain can learn continuously and from just a few examples. This research explores the impact of ’sleep’ an unsupervised phase incorporating stochastic network activation with local Hebbian learning rules on ANNs trained incrementally with limited and imbalanced datasets, specifically MNIST and Fashion MNIST. We discovered that introducing a sleep phase significantly enhanced accuracy in models trained with limited data. When a few tasks were trained sequentially, sleep replay not only rescued previously learned information that had been forgotten following new task training but also often enhanced performance in prior tasks, especially those trained with limited data. This study highlights the multifaceted role of sleep replay in augmenting learning efficiency and facilitating continual learning in ANNs.
more »
« less
- Award ID(s):
- 2223839
- PAR ID:
- 10544257
- Publisher / Repository:
- IEEE
- Date Published:
- ISBN:
- 979-8-3503-5931-2
- Page Range / eLocation ID:
- 1 to 10
- Format(s):
- Medium: X
- Location:
- Yokohama, Japan
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abstract Artificial neural networks are known to suffer from catastrophic forgetting: when learning multiple tasks sequentially, they perform well on the most recent task at the expense of previously learned tasks. In the brain, sleep is known to play an important role in incremental learning by replaying recent and old conflicting memory traces. Here we tested the hypothesis that implementing a sleep-like phase in artificial neural networks can protect old memories during new training and alleviate catastrophic forgetting. Sleep was implemented as off-line training with local unsupervised Hebbian plasticity rules and noisy input. In an incremental learning framework, sleep was able to recover old tasks that were otherwise forgotten. Previously learned memories were replayed spontaneously during sleep, forming unique representations for each class of inputs. Representational sparseness and neuronal activity corresponding to the old tasks increased while new task related activity decreased. The study suggests that spontaneous replay simulating sleep-like dynamics can alleviate catastrophic forgetting in artificial neural networks.more » « less
-
Bush, Daniel (Ed.)Artificial neural networks overwrite previously learned tasks when trained sequentially, a phenomenon known as catastrophic forgetting. In contrast, the brain learns continuously, and typically learns best when new training is interleaved with periods of sleep for memory consolidation. Here we used spiking network to study mechanisms behind catastrophic forgetting and the role of sleep in preventing it. The network could be trained to learn a complex foraging task but exhibited catastrophic forgetting when trained sequentially on different tasks. In synaptic weight space, new task training moved the synaptic weight configuration away from the manifold representing old task leading to forgetting. Interleaving new task training with periods of off-line reactivation, mimicking biological sleep, mitigated catastrophic forgetting by constraining the network synaptic weight state to the previously learned manifold, while allowing the weight configuration to converge towards the intersection of the manifolds representing old and new tasks. The study reveals a possible strategy of synaptic weights dynamics the brain applies during sleep to prevent forgetting and optimize learning.more » « less
-
Humans and most animals can learn new tasks without forgetting old ones. However, training artificial neural networks (ANNs) on new tasks typically causes them to forget previously learned tasks. This phenomenon is the result of “catastrophic forgetting,” in which training an ANN disrupts connection weights that were important for solving previous tasks, degrading task performance. Several recent studies have proposed methods to stabilize connection weights of ANNs that are deemed most important for solving a task, which helps alleviate catastrophic forgetting. Here, drawing inspiration from algorithms that are believed to be implemented in vivo, we propose a complementary method: adding a context-dependent gating signal, such that only sparse, mostly nonoverlapping patterns of units are active for any one task. This method is easy to implement, requires little computational overhead, and allows ANNs to maintain high performance across large numbers of sequentially presented tasks, particularly when combined with weight stabilization. We show that this method works for both feedforward and recurrent network architectures, trained using either supervised or reinforcement-based learning. This suggests that using multiple, complementary methods, akin to what is believed to occur in the brain, can be a highly effective strategy to support continual learning.more » « less
-
Abstract We present single‐column gravity wave parameterizations (GWPs) that use machine learning to emulate non‐orographic gravity wave (GW) drag and demonstrate their ability to generalize out‐of‐sample. A set of artificial neural networks (ANNs) are trained to emulate the momentum forcing from a conventional GWP in an idealized climate model, given only one view of the annual cycle and one phase of the Quasi‐Biennial Oscillation (QBO). We investigate the sensitivity of offline and online performance to the choice of input variables and complexity of the ANN. When coupled with the model, moderately complex ANNs accurately generate full cycles of the QBO. When the model is forced with enhanced CO2, its climate response with the ANN matches that generated with the physics‐based GWP. That ANNs can accurately emulate an existing scheme and generalize to new regimes given limited data suggests the potential for developing GWPs from observational estimates of GW momentum transport.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1109/IJCNN60899.2024.10650116
