Risk Analytics2018-09-18T08:42:47+01:00

Risk Analytics

Business Analytics and Machine Learning solutions for the analysis and detection of risks, fraud, financial scorings, delinquency, etc.


Fraud Detection

Fraudulent claims are one of the biggest problems for insurance companies and their detection can be of great benefit to companies.
In DatKnoSys we help you to detect those small claims using association, segmentation and predictive techniques, with which we look for anomalies that obey a certain pattern to analyze them in greater depth and detect possible frauds.


Portfolio management

DatKnoSys calculates the risk of a client’s different financial products, including the effects of diversification. Through predictive techniques we approximate the profit or price of a given financial instrument, and using optimization algorithms we focus our capital on activities that maximize profit or minimize risk as appropriate. Thus, depending on the requirements you request, we will make the portfolio recommendations that best suit your interests.


Estimate claim costs

From DatKnoSys we help companies to estimate the cost of a claim so that they can improve their forecasts and avoid the inconveniences caused by the extension of the agreements and the final magnitude that these claims can reach. This estimate will depend on factors such as the importance of the claim, the time it may take to reach an agreement, inflation and interest, etc …

The money reserved as a provision of funds for this type of event is immobilized until the end of the claim process, with which it is important to specify these costs in advance. Using data mining techniques, we created a predictive model based on the resolution of past claims, which gives us an estimate of the final amount of the claim, the time it is resolved and the amount that should be allocated to the provision of funds.


Analysis of Customer and Financial Market Risks

DatKnoSys calculates the risk of the different credits through data mining techniques that determine the risk associated with a credit based on the client’s history, socio-demographic variables, etc. For this purpose, we use predictive algorithms that classify the client in a specific risk category and, based on this assignment, we know how high the credit default risk is or if it requires a more in-depth study before being granted.

We also calculate the risk of the financial market through data mining techniques that allow us to develop models that measure the risk of different financial instruments based on indicators such as: interest rates, stock market indices or economic development. By means of classification or prediction models we assign a certain risk to a specific product and in this way we facilitate decision making.


Credits and Policies

DatKnoSys identifies the risk factors to establish the price of a premium, thus predicting possible claims and losses. These factors, obvious in some cases – a person who lives in the city is more at risk of a traffic accident than one who lives in the countryside, for example -, other times they are not that obvious and there may be relationships between difficult or impossible variables to identify without advanced techniques.

With a data mining model we predict the risk in a much more precise way, which allows us to better adjust the prices of the premiums and, consequently, reduce costs and increase profits. To build these predictive models, clients are segmented into homogeneous groups according to the purpose we want to achieve: identifying risk factors, behavior, benefit … Studying these segments we will obtain information on the characteristics of the groups that will help us determine the risks or behaviors of their components. At the same time, predictive models are applied to these segments that help us obtain more information about the behavior of each group.

Grupo CMC