Hello, as I said on my personal trading journal found here: ( https://www.forexfactory.com/showthread.php?t=627863 ) I am developing a new stat-arb strategy. I will be focusing on finding patterns in tick volume in order to track and predict large flows of capital between currencies. I thought it would be appropriate to form a discussion on predatory high frequency trading. I myself am not a market maker/scalper, due to the broker I use, but if you are a market maker/scalper feel free to discuss how your trading is going/provide personal insight.
I am exploring the use of neural networks when finding patterns in data. I am currently sampling and modeling data on 5m bars, but trading on 333ms increments. The reason for rounding to 5m bars is computational. It simply takes to much computing to do tick data. One day I hope to try tick data, but it's likely I will need to lease a cloud based vm. I've used cheap vms in the past to scrape and record tick data, but it can take 20-60 secs to backtest only 16 hours of tick data. This makes it impractical for neural networks.
I'm doing some visualizations of the data as it researches, so i can get a better idea of what some of the relationships plot out looking like. Here is an example.
I am exploring the use of neural networks when finding patterns in data. I am currently sampling and modeling data on 5m bars, but trading on 333ms increments. The reason for rounding to 5m bars is computational. It simply takes to much computing to do tick data. One day I hope to try tick data, but it's likely I will need to lease a cloud based vm. I've used cheap vms in the past to scrape and record tick data, but it can take 20-60 secs to backtest only 16 hours of tick data. This makes it impractical for neural networks.
I'm doing some visualizations of the data as it researches, so i can get a better idea of what some of the relationships plot out looking like. Here is an example.
Attached Image
Inserted Video
Quant Trader - My Blog: quantstop.blogspot (dot) com