In plain words: These models read a stock exchange's recent buy and sell orders and output a path of future prices, not one prediction. They match the usual single-prediction approach near-term but beat it further out, and train faster on a new AI chip than on GPUs.
Abstract · Multi-Horizon Forecasting for Limit Order Books: Novel Deep Learning Approaches and Hardware Acceleration using Intelligent Processing Units
We design multi-horizon forecasting models for limit order book (LOB) data by using deep learning techniques. Unlike standard structures where a single prediction is made, we adopt encoder-decoder models with sequence-to-sequence and Attention mechanisms to generate a forecasting path. Our methods achieve comparable performance to state-of-art algorithms at short prediction horizons. Importantly, they outperform when generating predictions over long horizons by leveraging the multi-horizon setup. Given that encoder-decoder models rely on recurrent neural layers, they generally suffer from slow training processes. To remedy this, we experiment with utilising novel hardware, so-called Intelligent Processing Units (IPUs) produced by Graphcore. IPUs are specifically designed for machine intelligence workload with the aim to speed up the computation process. We show that in our setup this leads to significantly faster training times when compared to training models with GPUs.
Zihao Zhang, Stefan Zohren
arXiv:2105.10430 · cs.LG, cs.NE, q-fin.TR · submitted May 21, 2021 · updated Aug 27, 2021
abstract · pdf · html · 18 pages, 7 figures, and 7 tables
With specialized hardware you can get close. But you’re still talking about a mid-single digit number of microseconds on inference alone. The competitor using linear models can get down to hundreds of nanoseconds. If you’re in FPGA world, that kind of latency advantage is worth way more than a 30% accuracy improvement from using a complex ML model.