Karol Przanowski

Background

About Karol Przanowski

My greatest passion is Credit Scoring (not necessarily used in banking). For the last 20 years I have been building, monitoring and validating various models: application, behavioural and debt collection for risk purposes and propensity models for cross-sell. I think scorecards, especially the simple WoE approach, are the best and the simplest tools for process optimization in case of lack of advance analytics before whereus AI techniques like XGBoosting in case of significant experience and trust to AI/ML models. Consequently, a large amount of money can be generated in any business process. My favourite theorem is: if the power of discrimination measured by Gini value is increased by 5% of points, then the profit of the credit acceptance process can be increased monthly by about 1,500 kPLN (300 kGBP, 500 kUSD, 350 kEUR). So I feel fairly confident in guaranteeing an increase (about 5%) in the predictive power in any credit acceptance process in any retail portfolio bank. Any costs involved in such a service are more than outweighed by the returns. In my book I present similar business cases for many different processes, typically connected with banking environment and in general for useful processes in any financial institution. My second passion is SAS, I am able to make calculations and analyses with SAS, especially with simple units such as: Base SAS, SAS/GRAPH and SAS/STAT. Macro-programming is my particular strength, and I am able to prepare fully automatised analyses in a very short time. Today I also use Python for coding, so I can accept the environment without SAS. I am aware of open-source revolution today and I share my passion also for Python. The third point that I focus on is the construction of fully automated advance processes, I can automatise most areas if they are connected with large processes such as large numbers of customers, large numbers of events etc. To summarise, I consider myself to be both creative and patient and capable of constructing SAS 4GL or Python coding. The ability to make an advance analysis while, at the same time, understanding the business environment in which I work makes me a Data Scientist who is well-suited for the Big Data revolution that is on the horizon. My documents: http://administracja.sgh.waw.pl/pl/OW/publikacje/Documents/Credit_Scoringstudia_przypadkow_Przanowski.pdf http://administracja.sgh.waw.pl/pl/OW/publikacje/Documents/ostateczny_CreditScoring_KPrzanowski.pdf http://arxiv.org/find/q-fin/1/au:+Przanowski_K/0/1/0/all/0/1