Fraud Detection

Data preprocessing techniques for detection, validation, error correction, and filling up of missing or incorrect data.
    • Calculation of various statistical parameters such as averages, quantiles, performance metrics, probability distributions, and so on. For example, the averages may include average length of call, average number of calls per month and average delays in bill payment.
    • Models and probability distributions of various business activities either in terms of various parameters or probability distributions.
    • Computing user profiles.
    • Time-series analysis of time-dependent data.
    • Clustering and classification to find patterns and associations among groups of data.
    • Matching algorithms to detect anomalies in the behavior of transactions or users as compared to previously known models and profiles. Techniques are also needed to eliminate false alarms, estimate risks, and predict future of current transactions or users.


Let you get more details from :


For banking


For insurance