Description: A summary of the Global Trends 2030 report, the fifth such report produced every four years by the National Intelligence Council. The report is based on the research and analysis of the 17 major intelligence agencies in the US. The summary covers the highlights of the report.

Description: Meet “Julia,” a free programming language licensed by MIT that may help you with parallel computing. It may be an alternative tool for those who are interested in nested stochastic processes for actuarial research (if not for regulatory compliance).

Description: This article describes the history and in some respects, the future of Big Data, and the tools we can use to handle it for PM and for machine learning. One such tool is Vowpal Wabbit, which sounds like something from a Bugs Bunny and Elmer Fudd cartoon (“Dwat that wabbit”), but it really is a popular approach to process a dataset of any size, as there is no need to load all the data into memory. Big data is forcing us to enhance many of our tools. So is the increased actuarial usage of stochastic-in-stochastic analyses (nested stochastic processes), which can result in unacceptably long program run times. Many of our popular computer languages are not inherently well suited for parallel computations.

Description: The author explains how he used judgmental forecasting techniques to develop a dynamic lapse assumption. He defines the techniques and provides readers with a step-by-step process.

Description: This paper explores how health care complexity challenges us to expand our repertoire of actuarial methods—and even recast our trend driver paradigm—in order to do our jobs better and maintain our credibility.

Description: In this continuation of our July 2013 article we explore an algorithm for evaluating Hidden Markov Models in detail. A preferred method in speech recognition technology, Hidden Markov Models are a promising technique when dealing with data you suspect may be driven by a change in a unobservable state. The example in the article is provided in an Excel workbook for the reader to follow along with.

Description: Deep learning is a method that allows neural networks to be constructed with more than just one or two hidden layers. These layers can then be trained independently and simultaneously of each other. This article provides an introduction to deep belief neural networks.

Description: The article introduces the concept of overfitting in predictive modeling and illustrates how it hinders our ability to predict new data well.

Description: This article discusses steps to follow to ensure a more rational and productive approach to data analysis.