Trading Catechism

H∞

There is a reason why EE majors are highly sought out in HFT.

 

Today, the latency is measured in nanoseconds.

 

Don't reivent the wheel. A very useful library is

https://cran.rstudio.com/web/packages/PerformanceAnalytics/index.html

But how do we use it from a friendly environment like c#?

http://rdotnet.codeplex.com/

 

Visualizing Machine Learning on streaming data with Spark:

 

Imo useful to go over each video as the subject matter interests you:

http://bigdatauniversity.com/

 

"Computer Science is no more about computers than astronomy is about telescopes" - Edsger Dijkstra

 

Sorry I missed this [interesting question!]

"...The implications of this are many. First and foremost the general property of e seems to be to split a complex function into its component, out-of-phase vectors. In this case, it does by factoring 1 and Phi into separate functions (notice that both integer (B) portion and Phi portion (A) are each multiplied by their own summations beginning with out-of-phase generations). Obviously e also does this in Euler’s Formula between the imaginary number i and 1. This is made evident by the following graphic:.."

http://vixra.org/pdf/1110.0045v1.pdf

BTW, notice in that graphic [Graphic 5] in that paper that what we see in nature is a projection of what is going on in a higher dimensional space - in my opinion in markets as well as in nature ("empty" space) they are spinors. That is almost certainly true in markets as well, what we see as price/volume playing out is a much higher dimensional action of some possibly infinite [Lie] group. It is the reason why eigen vectors/values are so important to physics and market models. When that projection is a Linear Lie Group, I believe technical analysis charting can work. But that is the exception, not the rule. A headfake is price action that is probably riding on a Mobius group.

In a sense, this could be taken as an atom unit of thought as to Ilinski's "markets are essentially electrodynamics":

http://arxiv.org/pdf/cond-mat/9806138.pdf

 

In essence, traditional technical analysis charting can work when the market is affording linear searches, in other words, it is in a pocket of inefficiency:

 

I will make another point. If charting is working, we are crossing a fixed point on the manifold. An extremely important regime change.

Imo, the best way to study regime change is in brain studies. So you study of regime change this way.

http://arxiv.org/abs/1402.6951