Save the date for our November R meetup. There will be four talks and afterwards the possibility for socialising.
Thanks to the speakers!
Drinks, fruits and snacks are sponsored by R Studio (https://www.rstudio.com/).
Make it simpleR! (Speaker: Raymond Zychowicz)
It is hard to present R as being simple. But its many advantages could be used to the benefits of many users outside the data analysis and statistical worlds. We are not only using R for our daily data health checks and support activities, but are also encouraging business analysts to make the first steps towards R. To achieve this, we must make it simpleR.
R in Finance: Using R in a productive and highly regulated environment (Speaker: Gabriel Foix)
In recent years R has become increasingly popular for data analytics and modelling in the finance sector. The power of R in statistical analyses is well known. However, in order to use R in a commercial and highly regulated environment, such as finance and insurance, new features and tool-kits were required. With important aspects like traceability, testability, auditability and scalability, R needs to be brought to the next level. Fortunately, the R world provides several tools to aid with these concerns, many of which not yet available a few years back, allowing us to build upon and helping us to integrate R into more complex company IT infrastructures. In this presentation, we will go through some of these tools using examples from the field.
R for Car Insurance Product (Speaker: Claudio Giancaterino)
One of the most useful example of R applied by Insurance Companies is the calculation of the car insurance premium (MTPL pricing). A non-life insurance policy is an agreement between an Insurance Company and the policyholder in which the duties of the Insurer is to cover the customer for certain unpredictable losses during a time period, usually one year, against a fee, the premium.By the insurance contract, economic risk is transferred from the policyholder to the Insurer. Due to the law of large numbers, the loss of the Insurance Company, being the sum of a large numbers of small independent losses, should be equal to its expected value. Furthermore the expected value of the losses can be defined by the product of the claim frequency and the claim severity distributions. For each policy, the premium is determined by the values of a number of variables: the rating factors (age, model of car, residential area…) and to estimate this relationship is employed a statistical model. Generalized Linear Models are the most popular statistical methods used in motor insurance to estimate rating factors; they are engaged, for a practical example, to estimate a commercial premium using an Australian dataset by one-way and multivariate approach. This approach is compared with Generalized Additive Models, used for continuous variables and with Generalized Non-Linear Models package.
Using Shiny for worldwide climate monitoring (Speaker: David Masson)
CelsiusPro is a technology company based in Zurich and specialized in structuring index-based insurance for weather, livestock, yield and natural catastrophes. In this talk, we will present a Shiny application used for climate monitoring worldwide. Specifically, we will show how large spatio-temporal fields such as climate gridded datasets can be analyzed 'on the fly' using MongoDB (non-SQL database) in a Shiny applications.