The rise of digitization leads to huge and massive data sets available for companies. A key challenge is to utilize those data sets for smart business decisions. In particular Machine Learning (ML) methods and Artificial Intelligence (AI) offer new opportunities to learn from data sets and gather new insights. This translates directly into a competitive advantage. Examples are improved dynamic pricing strategies and a better understanding of consumer behaviour based on state-of-the-art machine learning methods. We offer research-based state-of-the-art statistical and econometric methods for analysing your data sets. Moreover we have access to a network of scholars and scientists from the leading universities in the world which enables us to deliver unique and tailor-made solutions to our clients putting them ahead of the AI and Big Data era.
Economic AI Services
ML and AI
The use of modern methods (Machine Learning methods and Artificial Intelligence) to create value out of data for businesses.
Prediction Analytics
for economics and finance problems based on scalable ML / AI
Development and implementation
of pricing strategies to improve revenues and profits, development of rating based on ML methods in the insurance business
Double Machine Learning
Causal Inference and Programme Evaluation using Modern Methods
Data-driven Mechanism Design
Incentive schemes, Auction designs, Pricing schemes, Recommendation systems based on ML / AI
Experts opinion
by frontier academic researchers for unique problems
Short, Intensive Courses
for Analytics/Research & Development Professionals bringing the frontiers of ML for causal inference and intelligent decision making
Prediction Analytics
for economics and finance problems based on scalable ML / AI
Examples:
- Fraud Detection (Insurance, Finance)
- Claims Prediction (Insurance)
- Quitting behaviour and change of service provider of customers (Insurance, Banking)
- Demand estimation/estimation of elasticities
- Credit Scoring (Finance and Banking)
- Load Forecasting
- Operational Risks for Banks
Introduction
Dynamic pricing strategy
In this project a dynamic pricing scheme was implemented for a major Asian ride-sharing company that helped to maximize gross revenue. In the first step the demand function for rides was estimated from the company’s data set with Deep Learning, improving upon traditional estimation methods for demand. From this estimated demand functions novel insights were obtained, e.g. the influence of weather conditions or concave pricing in distance. The demand function was used as an input to “solve” for the optimal pricing policy function which maximize gross revenue. The proposed pricing system leads to a small price increase on average, but 40% of the price will be lower. In sum, revenues could be increased by more than 5% and profits by an even larger amount.
Fraud detection
In the private health insurance bills are often flawed leading to reduced payments to the health service providers and hence saves money for the insurance company and insurees. Within this projects different architectures for Deep Learning for this problem were designed, implemented and compared. The prediction rules based on Deep Learning out-performed other Machine Learning methods and achieved very good results leading to better management of claims in the health insurance.
Operational Risk
Computing Economic Capital requirements for banks using robust, rigorous methodology for fitting loss distributions that accurately describe most common loss types occurring in banking. The methods have been successfully deployed at a major U.S. banking corporation.
About us
The company comprises a network of partners, who are top-level researchers and practitioners and leaders in the field of econometrics, machine learning, mathematical statistics, and operations research. They work as professors at Harvard University, University of Hamburg, Massachusetts Institute of Technology, University of Chicago, Hong Kong University, New York University, and Duke University. The partners have extensive publication records and are winners of major academic awards and have completed projects with Microsoft, Amazon.com, State Street Corporation, Tata, DIDI, and others.
Contact
Economic AI GmbH
Nürnberger Str. 262 A
93059 Regensburg
Germany
eMail: info@economicai.com
Tel.: +49 9401 92080
Fax: +49 9401 8336
Director: Prof. Dr. Martin Spindler