Develop a risk management system to track fraudalent transactions.

Hi! I'm an engineer in statistics and applied economics. In fact, I have a relevent experience on the credit risk Management. So I suggest you to conduct a classification that w'll permit to distinguish between the fraudulent and non fraudulent clients, than we can explain this phenomena using a logit regression model and finally provide an accurate predictions for the new clients . Be sure that i can provide you a clean report with rigorous statistical analysis and within the beefiest delay.... Get your statistical job done ... Perfectly

We are a team of experts which are dealing with various projects such as writing, technical writing, Engineering, PCB designing, FPGA, Verilog /VHDL, MATLAB, Mathematics, Calculus, SPSS, Statistic, CUDA, OpenGL, Pattern recognition, Image processing, signal processing, C++/C programming, ETC. We are ready for hiring right now.

We need to build a Fraud detection model - classify sales as fraudulent or not based on historical data. Algorithms like Decision Tree, Random Forest, Logistic Regression are used. Need to look into sample data and arrive at the right algorithm that fits the data and the nature of business. Once built, the model can be deployed to classify new transactions on a real-time basis. It can be made available as a REST API that can be called by front-end systems before processing a transaction. PS: I'm not from US but can work for a good overlap with your timezone to make sure the project runs smooth.

************Chat me now please if you chat me i can help you very low price****************** *************************************************

Dear Valued Employer, I can add value to your project, by suggesting new ideas. I am fast and reliable. For further understanding why I am the right firm for your project, kindly check my profile, reviews, rating and portfolio. With over 8 years of cross-functional experience, and helping clients world wide with diversify team, I provide genuine and innovative solutions as per your circumstances and requirements. I am MBA-Finance, M.A-Economics, and Qualified Chartered Accountant. If you feel to discuss more before you award me the project, you can send me the message. Kind regards, Rakesh

Hello Sir/Madam, hope you are in pink health. I have worked as a Fraud Analyst for nearly 2 years for a top travel website based in USA which also happens to be a Fortune 500 company. I performed pretty well through out my tenure and was regularly a top performer after the probation period of 3 months. My job responsibility was to scrutinize the reservations or transactions on a real time basis and to determine whether to allow a transaction or decline it. I was one of the team members who was looked upon to find the patterns to find out the fraudulent reservation on the real time attempts and declining it immediately and hence preventing fraud. Coming to the part why you should hire me. I usually saved $30000 in a month which eventually saved the client $50000 in a month and I did it with a negligible chargeback and false positive hit. To design a system that tracks suspicious transactions definitely programming is required but more importantly we need rules or logic which is to be applied. For example let us say that the IP city is 500 miles away then the billing city then it can trip the rule which won't decline the transaction but will bring the transaction as suspicious and if there are many such rules that apply on a transaction then we may even decline it. So, I suggest you here that you need a blend of two knowledge base, one who can code and the other who knows what need to be coded.

Hi, Im a PHD student in Computer Science currently Studying in RPI which is in Troy, NY. Data analytics is my research field and I code in R for my research every day. Therefore, I think my skills are perfectly compatible with your requirements. I would need more details regarding what kind of system you are actually looking for and also the kind of data you actually have for the system. Typical fraud detection systems require prior data with clear classification labels such as fraud/safe. These kinds of data could be modeled with a supervised or classification algorithm. But unlabeled data can also be modeled given a strong set of data about the transactions themselves. There are also many other factors that will dictate the choice of modeling algorithm such as whether the model in the system will be online (continuously updated), sensitivity for false positives and so on. So, please let me know more details if you are interested.