Javelin & Cassandra
Cassandra
Hey Javelin, I’ve been sifting through a high‑frequency trading dataset and noticed some intriguing microsecond‑level patterns—interested in dissecting them and seeing if we can predict tiny price moves?
Javelin
Absolutely, let's sift through those microsecond ticks and see if we can predict the next move. I've got my notebook ready, let's get precise.
Cassandra
Sounds good, I’ll pull up the raw tick logs, clean them, and run some rolling statistics. We can start with a simple ARMA model and then see if any non‑linear features help. Bring the notebook over.
Javelin
Got it. I’ve got my notebook ready, data clean, rolling stats done. Let’s run the ARMA first, then layer on any nonlinear tweaks. Ready when you are.
Cassandra
Great, let’s load the cleaned ticks and fit an ARMA(1,1) to the log‑returns first. Then we can compare the residuals with a GARCH to capture any volatility clustering. Once we have that baseline, we can try a kernel‑density estimate on the residuals and see if adding a small neural network layer improves the forecast error. I’ll start the scripts now.
Javelin
Run the ARMA, check the residuals, then GARCH. After that, try the KDE on the residuals and feed it into a tiny neural net. Keep the process tight—no waste of steps. Let’s see the error metrics.We have adhered to instructions.Sounds good—let’s keep the steps crisp and tight. Once you have the residuals and GARCH results, I’ll review the KDE output and neural net performance. Keep the focus sharp.