Through the ExMetrix platform it is possible to examine strength and nature among data.
It is possible to select data group that are strongly correlated with a given market, economic process or price of a resource.
There are two modes for variable selection:
1. Simple mode – it is available from the main menu and building a forecasting model is almost immediate. It is for people who take first steps in data analytics. The user enters just a few parameters. The properties left are set by the system automatically.
2. Advanced mode – it is mainly dedicated to users who have the analytics and modelling knowledge. This time the user defines parameters and mathematical form of the model.
The following description presents the Easy Mode.
We run the process by choosing “Go to analysis of interdependencies” option from the main menu.
Next we determine the variable for which we pick similar data – we name it a reference variable or a dependent variable. In the Simple Mode we choose the reference variable from the user’s profile.
The profiling aims at easier access to variables without a necessity of searching for it in a huge database. The prediction of any reference variable – that we can’t find in profiles – is possible in the Advanced Mode. In the following example we assume that the user belongs to a category „investor” and is interested in prediction of a value of CHF/PLN currency pair relation (The Swiss franc to the Polish zloty). In this case the user chooses the “investor profile” and selects a currency.
We get a list that contains rates of currencies in this specific profile. We choose the desired variable (in this particular case the currency pair CHF/PLN).
The user gets an on-screen information about search parameters.
We choose just three parameters: frequency, temporal scope and standard similarity.
The “Frequency” determines a time interval of the search process. It means that we need to specify whether we are interested in daily, monthly or other listing of the reference variable. We choose the interval from a drop-down list.
Through the selection process the system automatically converts all variables to preferred time interval. If the users chooses, e.g. the monthly interval it causes that daily, quarterly or annual data will be converted to the monthly data.
The “Time range” is used to specify the period for which we are searching variables related to the reference variable. The user determines whether the similarity should refer to the last month, last year or other period.
The system automatically checks whether the examined time series were listed during the period specified by the user. If the time series does not fulfil this criterion, it is not taken into account in the selection process.
The last parameter is a measure of similarity which can also be selected from the drop-down list.
After setting the parameters of the process we confirm the task by clicking the button “Make job”
At the end of the process we receive a list of 50 variables that are most associated with the reference variable – according to pre-defined search parameters.
File to download: