Return to main page
Detailed Solutions to Difficult Problems
Print this Article  Send to Colleague
While many actuaries pride themselves on the depth and extent of their expertise in the world of probability and statistics, there may be times when even their knowledge comes up short.
Fortunately, there is an internet community,
Stack Exchange, specifically designed to connect those with problems and those with possible solutions. It is composed of 89 sites, which include at least three that may be of value to CIA members.
Cross Validated describes itself as "a free, community driven Q&A for statisticians, data analysts, data miners and data visualization experts", and its message board is designed to help those working on such subjects as:
 Statistical analysis, applied or theoretical;
 Designing experiments;
 Collecting data;
 Data mining;
 Probability theory;
 Mathematical statistics; and
 Statistical and datadriven computing.
Actuaries are among its varied members, who are given a reputation rating based on the expertise they demonstrate in their answers to other people's questions. The most highlyrated contributors include holders of PhDs and master's degrees in statistics; professors; and many more experts in the field.
Visitors, who do not have to register, are encouraged to pose practical, answerable questions based on actual problems that they face. Recent questions involve such topics as the ratio between two densities, detecting outliers, interpreting forms of a mixed model, logistic regression, and cluster analysis.
Actuaries with questions that are more mathfocused will find a similar high level of support at the Mathematics section of the Stack Exchange network. There, a retired professor emeritus and several PhDs are among the the toprated users, and the topics under discussion include duality, functional analysis, convergence, algorithms, logical expressions, and graph products.
Another site in the network which may prove invaluable for CIA members is
Quantitative Finance, where users look at topics like derivatives, the difference between kappa and delta, variance swaps, and spectrum analysis. Thirty questions have been asked about modelling alone, and another 20 about risk management.
Pros:
the detail and expertise provided by users; the variety of questions.
Cons:
despite the level of knowledge available, some questions do go unanswered; the layout is not as attractive as some other sites.
Addresses:
Stack Exchange http://stackexchange.com
Cross Validated http://stats.stackexchange.com
Mathematics http://math.stackexchange.com
Quantitative Finance http://quant.stackexchange.com

