Saturday, August 2, 2014

Time to create an integrated model to predict "2012" or "Die Hard 4"

Rare event is an event that happens very less in our day to day life. For instance, the event that we all saw in the film "2012" or in "Day After Tomorrow" Motion picture can be termed as a rare event. Thus the melting and displacement of earth's crust, natural disasters comes under the domain of rare event. The 2007 motion picture flip "Die Hard 4" also had a different form of rare event. It showed us about an organized terror attack that was coupled with cyber and network security attacks creating a complete shutdown of road, power, banking infrastructure. The commonality between all these events is that they are rare but have dire consequences and cannot be predicted easily as the chances of estimating the occurences of these events are often beyond the imagination of mathematicians, modellers. However, they are still possible and they do happen with high damages. Moreover when they happen if the preparation to manage those damages are less the risk of damages to society goes up. Therefore, from a perspective of risk management, it is important to know the chance of occurrence of these rare events so that measures can be taken to manage them. Risk mitigation in case of rare events should also be tried in case of some types of rare events like the terror attack mentioned in case of the film Die Hard 4. It can reduce the extent of damages and uncertainty (simply put as the consequences which are beyond the control of risk mitigation measures at a point of time) associated with the occurence of the event. However, for rare events like as depicted in the film 2012: risk mitigation measures can be difficult as controlling the occurences of the event was totally outside the human control (though the team knew it in the film few months in advance that something will happen) and was a natural phenomenon.


Therefore, it is important to understand the pattern of occurence of these events even though it is difficult to put them in a statistical distribution pattern. However, through use of multiple simulations by using Monte Carlo techniques (which creates an imaginary world of simulation and runs such a simulation again and again) and uncertainty modelling if the chance of their occurrence can be predicted even with a margin of error, it can prepare the human society to prepare in a better way to create risk management measures through preventive actions. Policies can be therefore designed on the basis of these predictions (considering the margin of errors in the predictions) to create tangible measures that can protect the human society from adverse consequences of these events when they happen. However, rare event modelling as of now is often not done in an integrated manner. Time has come to join the skills of engineers, psychologists, scientists, modellers, social scientists, people from different disciplines and policy makers to establish integrated rare event models which will capture all the elements of consequences of a damage that can happen to the human society in case of occurrence of a rare event.



For rare events, which do not have well defined probability of occurence, in other words for whom the uncertainty of probability estimates are very high, there is a larger need for creating integrated rare event models. It is because for most of these events, the damage extent and degree will be very high. Therefore the importance of mitigating risks in such cases assumes a larger relevance. Mitigation of such risks can only happen when within the construction process of a rare event model, there is an engineer who can say that what can be the chances of failure of certain engineering designs of an infrastructure in the wake of a rare event (say a natural disaster). It will also require a scientist in the model building who will incorporate the science of the natural process owing to which the natural disaster happens. Behavioral, social scientists, psychologists will need to pitch in the model-building by suggesting how people can behave in the wake of such a rare event. In case of occurrence of a human induced rare event of an extreme type like cyber and infrastructure network terrorism (as shown in Bruce Willis starrer Die Hard 4), psychologists have to be a part of the modelling process to understand the mind pattern of the terrorists who aim to attack cyber and infrastructure networks. Such an understanding will help in framing some surrogate probability distribution pattern to predict these events for possible risk mitigation measures. Finally, within the modelling team  policy makers, practitioners, have to be there who can constantly question the modelling team to be sensitive to the practical domains of policy realities coming out of the ideal world of modelling simulations while framing the equation structures and finalizing the result interpretations. This will help in coming up with policy prescriptions to create risk mitigation measures for addressing the consequences of rare event occurences. Once, all this integration happens, rare event modelling work will transform to an useful tool to capture the realistic world of imaginations that can happen in future and need to be tackled to protect the human society. However, to achieve that, rare event modelling has to be more of an art based on the science of understanding an imaginary world of occurence of certain events being placed in the present context. Therefore, a good piece of rare event modelling is one where there are many scientists, thinkers, implementors who have turned as artists and are constantly working together to move inwards towards different imaginary layers of dreams as was shown in the motion picture -  "Inception".