Using Natural Language Processing and Reinforcement Learning to predict stock prices from google news feed using techniques from many state of the art papers
Longformer (Sliding window attention to scale compute linearly with sequence length)
https://arxiv.org/pdf/2004.05150.pdf
Big Bird (Window attention combined with random attention)
https://arxiv.org/pdf/2007.14062.pdf
Transformer XL (Using cached query and keys for extra long sequences)
https://arxiv.org/pdf/1901.02860.pdf
Seed RL (Scalable and Efficient Deep RL with Centralized Inference):
https://arxiv.org/abs/1910.06591
R2D2 (Rainbow Deep Q Network with LSTM):
https://openreview.net/pdf?id=r1lyTjAqYX
PCGrad (Gradient Surgery for Multi Task Learning):
https://arxiv.org/pdf/2001.06782.pdf
Block Recurrent Transformer (Variant of Transformer with Recurrent Attention):
https://arxiv.org/abs/2203.07852
Implicit Quantile Network (Distributed RL for accelerated learning and risk sensitive policies):