System Monitoring
Finding correlations with COR-RiSTOR Software as a Service
The real estate market is a market where suppliers and buyers of real estate meet. If supply and demand are balanced, there is market equilibrium and the prices are equilibrium prices (Stagnating). Market imbalance exist when there is a supply gap. Then the market prices are Inreasing. Market imbalance also exist when there is a demand gap. Then the market prices are Falling. The real estate market has specific characteristics. It strongly depends on the Inflation and on the Key interest rate.
As an example, we take the Geman real estate index (GREIX) data and predict whether the GREIX is Increasing, Stagnating, or Falling. Using clustering of GREIX data after processing with artificial neural networks, it is difficult to separate Increasing GREIX from Falling GREIX and to separate Stagnating GREIX from Falling GREIX. The obtained recognition rate is only 51%. Using clustering of GREIX data after preprocessing with the COR-RiSTOR algorithm we obtain a recognition rate of 82 %.
Min(Inflation) = -0.6
Max(Inflation)=+8.8
Min (Key interest rate) = 0
Max(Key interest rate)=+4.5
Min (GREIX) = 85.89
Max(GREIX)=438.05
System monitoring: Preprocessing real estate market data with the COR-RiSTOR algorithm (SaaS)
Clustering of real estate market data after preprocessing with artificial neural networks
Clustering of real estate market data after preprocessing with COR-RiSTOR algorithm
Using the software as a service (SaaS) we determined a nonlinear correlation between Data A (Inflation) and Data B (Key interest rate), a linear correlation between Data A (Inflation) and Data C (GREIX), and a linear correlation between Data B (Key interest rate) and Data C (GREIX). It is mainly the non-linear correlation between Data A and Data C which supports the separation of Increasing GREIX from Falling GREIX and to separate Stagnating GREIX from Falling GREIX.
Preprocessing with COR-RiSTOR
Memristor 1 (Data A&B)
Memristor 2 (Data A&C)
Memristor 3 (Data B&C)
