Transformers as Statisticians: Provable In-Context Learning with In-Context Algorithm Selection.
Yu BaiFan ChenHuan WangCaiming XiongSong MeiPublished in: NeurIPS (2023)
Keyphrases
- learning algorithm
- selection algorithm
- dynamic programming
- learning speed
- detection algorithm
- k means
- optimization algorithm
- worst case
- computational complexity
- context aware
- cost function
- np hard
- expectation maximization
- experimental evaluation
- improved algorithm
- matching algorithm
- contextual information
- incremental learning
- learning phase
- learning rules
- significant improvement
- objective function
- learning process
- preprocessing
- learning problems
- simulated annealing
- classification algorithm
- theoretical analysis
- input data
- neural network
- high accuracy
- reinforcement learning
- decision trees
- genetic algorithm