Contextual Memory Trees
arXiv:1807.06473
Abstract
We design and study a Contextual Memory Tree (CMT), a learning memory controller that inserts new memories into an experience store of unbounded size. It is designed to efficiently query for memories from that store, supporting logarithmic time insertion and retrieval operations. Hence CMT can be integrated into existing statistical learning algorithms as an augmented memory unit without substantially increasing training and inference computation. Furthermore CMT operates as a reduction to classification, allowing it to benefit from advances in representation or architecture. We demonstrate the efficacy of CMT by augmenting existing multi-class and multi-label classification algorithms with CMT and observe statistical improvement. We also test CMT learning on several image-captioning tasks to demonstrate that it performs computationally better than a simple nearest neighbors memory system while benefitting from reward learning.
arXiv commentICM 2019
References in corpus (11)
- Very Deep Convolutional Networks for Large-Scale Image Recognition
- Scikit-learn: Machine Learning in Python
- Prototypical Networks for Few-shot Learning
- One-shot Learning with Memory-Augmented Neural Networks
- Learning to Remember Rare Events
- Conditional Probability Tree Estimation Analysis and Algorithms
- Hierarchical Memory Networks
- Logarithmic Time Online Multiclass prediction
- Search Engine Guided Non-Parametric Neural Machine Translation
- Learning Efficient Algorithms with Hierarchical Attentive Memory
- Logarithmic Time One-Against-Some