In data mining, modelling SOCIO uses the business process called CRISP-DM which guarantees at each stage of the project; documentation and portability.
To allow you to integrate our recommendations internally for operational and perennial uses of your models is one of our commitments.
A simple process
We focus on responding to your business issues. Teams who wan integrate your teams. We use standard market software as PASW Modeler (ex Clementine), SAS EM or Rapid Miner.
Third generation languages allow us to reintegrate economically our models in your own IT system.
Our fields of interventions
- Customers segmentation (SMB/FRM/FRAT)
- Customer value, loyalty and profitability models
- Determination of customer potential
- Various scorings : response, propensity to buy, churn
- Customer experience and profiling
- Database enrichment : Access bas method
Methods
- All regression techniques
- Decision trees
- Neural networks, Bayesian networks
- Automatic classifications
- SVM / Genetic algorithms
- Market basket analysis techniques
- Discrete choice models
- Forecasting : ARIMA / SARIMA
