[Okta Webinar] Learn how to a build a cloud-first strategyRegister Now

x
  • Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 1095
  • Last Modified:

Interpreting Curves - Rates of Change etc...

hi,

i apologise up front for the vagueness of this question!

i collect various data types(primarily odds from betting exchanges) and graph this data which all looks very pretty.
from the graphs i can visually identify trends etc... such as sudden/longer term movements
The problem i have is that i now need to develop an application that can interpret the graph and automatically identify the trends that are occurring.
I've looked into areas such as rates of change(slopes), differential calculus, moving averages etc... all in an attempt to find a way of defining the graph/curve at certain points and over a specified period.
admittedly this is not an area that i have any great experience in. so i find myself guessing a lot and am not sure where to focus in on to make any progress.

so the question i have is, as a general rule what areas should be focussed on when attempting to define a curve.
i realise that the answer to this question is dependant on the data being analysed, but i'm really looking for a starting point.
for example, am i better concentrating on rates of change over ceratin periods. what does finding the derivative at various points really tell me? a

any and all suggestions are welcome!
again, sorry about being so vague!

regards
0
Fatlog
Asked:
Fatlog
2 Solutions
 
NovaDenizenCommented:
What you're talking about sounds a lot like Technical Analysis, which is the term for a general art of analyzing stocks from chart information.  That would make a good starting point.  http://en.wikipedia.org/wiki/Technical_analysis
0
 
aburrCommented:
moving averages has a long history of use in predicting sunspot activity.
If you are looking for unusual betting activity, you might calculate the dirivitive of your curve and program an alert to sound when it changes by a predetermined amount.
0
Concerto's Cloud Advisory Services

Want to avoid the missteps to gaining all the benefits of the cloud? Learn more about the different assessment options from our Cloud Advisory team.

 
mandeliaCommented:
<<<as a general rule what areas should be focussed on when attempting to define a curve. >>

Derivatives.  Derivatives of a curve tell you a lot about the tind of curve.
The more deep you go. Degree of derivatives better you can mathematically define your curves
0
 
hilayCommented:
I agree with every thing written here and must say it similar to technical analyze of stock market and the same rules will work in your case. learn a little bit on the subject and get to know terms like signal, entry point, exit point, strategies and find the eqvilent in your field.

Mandelia - I think derivatives will not work there as it a statistical data that unlike "regular" function (of any order) would  work only on a given data segment without prediction of out-scope segments.
 
0
 
mandeliaCommented:
Hilay- I suggested derivatives because the question specifically said "Rate of Change".
Even if the data i collected samples like in a stock market. derivatives is what we use to analyse data.
The only difference is we segment our data. derivatives, local maximas and minimas, support and rate of growth are all some of the most common form of derivatives that a stock analyst looks at.
In stock analysis, we first need to smoothen data ignoring spikes and then analyse it. Even conditioning of data differs from one analysis to other
0
 
FatlogAuthor Commented:
hi,

thanks for all the comments!!! keep them coming.
i am very new to this and am trying to keep my head above the water at the moment so all comments help.
the one thing i need to be able to achieve is some way of automating the analysis of data so that a person isn't involved.

on another note, the markets i am looking at are odds markets that generally start every morning and finish sometime during the day. in general what would be a good time frame to analyse the data. seconds, minutes, hours?
also, these markets tend to become very volatile close to the market close time when large volumes of money are being fed into the market. is this type of sudden movement even predictable?

i appreciate that you may not know the markets in question and therefore are unable to give a concrete opinion. but i'm wondering if some of the behaviours that i see in these markets are common in other areas/markets. does volume in general dictate movement?

thanks again
0
 
mandeliaCommented:
Fatlog : Are you trying to develop such a analytical tool or you are open to buying these things.

If you just want a idea of what kind of analysis can be done you can look at matured analytical tools used for stock market analysis or the one used by google Analytics. The number of ways a data can be analysed depends on what you want to get from it. Rest everything can be automated.

Just an example : Farcs, Fanlines, Cyclelines, Parellels are few studies that can be done. These can be done on both volumes, odds, % move or any other dimention of the market. All these can be programmed to have programatically. But the harder question is to know what you want to see
0

Featured Post

Free Tool: Site Down Detector

Helpful to verify reports of your own downtime, or to double check a downed website you are trying to access.

One of a set of tools we are providing to everyone as a way of saying thank you for being a part of the community.

Tackle projects and never again get stuck behind a technical roadblock.
Join Now