mathcube.de Dr Radoslava Mirkov, data science - data analytics - statisticals, Berlin, Germany
Mentoring business leaders and managers understand and use data analytic tools on their own
Consulting of clients on topics like
Coaching and lecturing high school children in mathematics
S-Rating-Risikosysteme GmbH, Berlin, Germany
Design and implementation of score calculation engine a part of a tool for automated validation of scoring models compliant with regulatory requirements
Lendico Deutschland GmbH, Berlin, Germany
Design, development, implementation and validation of challenger credit scoring models
Forecasting of online sales
Coaching junior colleagues in data science
Assessment of currently used scoring models regarding model risk and audit readiness
Prodigy AI GmbH, Berlin, Germany
Design, development, implementation and validation of real-time machine learning models used for credit scoring based on various classification methods
Conception and implementation of APIs to embed models in the existing system design
Monitoring model performance, as well as investigating the factors that influence the performance, such as cut off strategy, marginal PD, retention rate, LTV, IRR, etc. on the cross-roads of risk and marketing
Design, development and implementation of financial and customer reports giving the overview of performance on daily, and monthly as well as ad hoc basis, used in management and board meetings for decision making
Project management of all data science and BI/data analytics projects including management of external service providers
Collaboration with C-level management to provide relevant data and predictive tools to support strategic decision making
Coaching colleagues from other departments to improve their analytic skills
ICE International Copyright Enterprise Germany GmbH, Berlin, Germany
Statistical analysis and modelling of demand, internal workload and productivity to support the development of the new price book
Design, development, implementation and automation of internal and customer KPIs, dashboards and reports giving the overview of performance metrics on daily, weekly, monthly as well as ad hoc basis, used for work allocation, customer, management and board meetings
Investigation of factors that influence important internal KPIs, such as automation rate, synergy effect and productivity
Collaborate with executive management to understand business needs and provide relevant data to support strategic decision making
Present reporting design and results to customers
Data architecture design to enable implementation of a professional BI solution, selection and evaluation of BI tools
Coaching new colleagues, involvement in hiring process for BI/Data Analytics team
TriSolutions GmbH, Management Consulting, Hamburg, Germany
Modelling and forecasting customer behaviour of credit taker based on multivariate data mining techniques, customer segmentation based on unsupervised machine learning methods
Validation of financial models for market, liquidity and operational risk as well as rating models
Pricing of financial products and testing the development of an ALM system
Workshops Model Validation and Quantitative Methods in Market Risk Management, conception, presentation of selected topics and hosting the discussions
In German, English and Serbian
For detailed list of topics and subjects, see the section Teaching in the menu
Humboldt Innovation GmbH and Humboldt-Universität zu Berlin, Department of Mathematics, Berlin, Germany
Working group of Prof. Dr. Werner Römisch, in project "ForNe - Research Cooperation Network Optimization", project lead Prof. Dr. Thorsten Koch, Technische Universität Berlin (http://www.zib.de/projects/forne-research-cooperation-network-optimization)
Development and implementation of algorithms for monitoring and forecasting gas loads as part of an online platform for capacity optimization of the largest gas network in Germany
Statistical data analysis and modelling by the means of nonlinear and semi-parametric methods for forecasting gas consumption at the design temperature, a special type of supervised machine learning methodology
Input parameters for scenario generation with an aim of optimization of gas networks
Presentation of results at workshops with the industrial partner and international conferences as well as in publications
First leadership experiences with students working on projects
Bank Austria UniCredit Group, Market Risk Management and Risk Analysis, Vienna, Austria
Development and roll-out of Market Conformity Information System, an online (intranet) tool for market conformity control
Statistical analysis of market data and its impact on risk measurement and reporting
University of Vienna, Faculty of Business, Economics and Statistics, Department of Statistics and Operations Research, Vienna, Austria
Under supervision of Prof. Dr. Georg Pflug
Dynamic stochastic programming including possible solution methods
Approximation of stochastic processes and error analysis
Joanneum Research GmbH, Department Statistical Data Analysis und Graz University of Technology, Department of Statistics, Graz, Austria
Under supervision of Prof. Dr. Herwig Friedl
Explorative data analysis for health studies
Statistical data analysis and forecasting based on generalized linear models for medical and environmental studies
S-Rating-Risikosysteme GmbH, Berlin, Germany
Design and implementation of score calculation engine a part of a tool for automated validation of scoring models compliant with regulatory requirements
Lendico Deutschland GmbH, Berlin, Germany
Design, development, implementation and validation of challenger credit scoring models
Forecasting of online sales
Coaching junior colleagues in data science
Assessment of currently used scoring models regarding model risk and audit readiness
Prodigy AI GmbH, Berlin, Germany
Design, development, implementation and validation of real-time machine learning models used for credit scoring based on various classification methods
Conception and implementation of APIs to embed models in the existing system design
Monitoring model performance, as well as investigating the factors that influence the performance, such as cut off strategy, marginal PD, retention rate, LTV, IRR, etc. on the cross-roads of risk and marketing
Design, development and implementation of financial and customer reports giving the overview of performance on daily, and monthly as well as ad hoc basis, used in management and board meetings for decision making
Project management of all data science and BI/data analytics projects including management of external service providers
Collaboration with C-level management to provide relevant data and predictive tools to support strategic decision making
Coaching colleagues from other departments to improve their analytic skills
ICE International Copyright Enterprise Germany GmbH, Berlin, Germany
Statistical analysis and modelling of demand, internal workload and productivity to support the development of the new price book
Design, development, implementation and automation of internal and customer KPIs, dashboards and reports giving the overview of performance metrics on daily, weekly, monthly as well as ad hoc basis, used for work allocation, customer, management and board meetings
Investigation of factors that influence important internal KPIs, such as automation rate, synergy effect and productivity
Collaborate with executive management to understand business needs and provide relevant data to support strategic decision making
Present reporting design and results to customers
Data architecture design to enable implementation of a professional BI solution, selection and evaluation of BI tools
Coaching new colleagues, involvement in hiring process for BI/Data Analytics team
mathcube Dr Radoslava Mirkov, education - data analytics - statistical modelling, Berlin, Germany
Valuation of swaptions in the current market environment
Validation of models for credit and liquidity risk
Prediction of leavers based on classification with survival analysis including recommendations of the measures to keep employees, by the means of the machine learning methods
Coaching and lecturing high school children in mathematics
TriSolutions GmbH, Management Consulting, Hamburg, Germany
Modelling and forecasting customer behaviour of credit taker based on multivariate data mining techniques, customer segmentation based on unsupervised machine learning methods
Validation of financial models for market, liquidity and operational risk as well as rating models
Pricing of financial products and testing the development of an ALM system
Workshops Model Validation and Quantitative Methods in Market Risk Management, conception, presentation of selected topics and hosting the discussions
In German, English and Serbian
For detailed list of topics and subjects, see the section Teaching in the menu
Humboldt Innovation GmbH and Humboldt-Universität zu Berlin, Department of Mathematics, Berlin, Germany
Working group of Prof. Dr. Werner Römisch, in project "ForNe - Research Cooperation Network Optimization", project lead Prof. Dr. Thorsten Koch, Technische Universität Berlin (http://www.zib.de/projects/forne-research-cooperation-network-optimization)
Development and implementation of algorithms for monitoring and forecasting gas loads as part of an online platform for capacity optimization of the largest gas network in Germany
Statistical data analysis and modelling by the means of nonlinear and semi-parametric methods for forecasting gas consumption at the design temperature, a special type of supervised machine learning methodology
Input parameters for scenario generation with an aim of optimization of gas networks
Presentation of results at workshops with the industrial partner and international conferences as well as in publications
First leadership experiences with students working on projects
Bank Austria UniCredit Group, Market Risk Management and Risk Analysis, Vienna, Austria
Development and roll-out of Market Conformity Information System, an online (intranet) tool for market conformity control
Statistical analysis of market data and its impact on risk measurement and reporting
University of Vienna, Faculty of Business, Economics and Statistics, Department of Statistics and Operations Research, Vienna, Austria
Under supervision of Prof. Dr. Georg Pflug
Dynamic stochastic programming including possible solution methods
Approximation of stochastic processes and error analysis
Joanneum Research GmbH, Department Statistical Data Analysis und Graz University of Technology, Department of Statistics, Graz, Austria
Under supervision of Prof. Dr. Herwig Friedl
Explorative data analysis for health studies
Statistical data analysis and forecasting based on generalized linear models for medical and environmental studies