Swine Flu or Anti-immigration

The following article talks about why US is not closing borders with Mexico. I guess the fundamental objective of closing the border is to contain the virus. However when the virus has already spread to US, closing borders will have no impact.

http://www.time.com/time/health/article/0,8599,1894786,00.html?imw=Y

Simulation models state that even the strictest limits on air travel would only slow the start of a flu pandemic, not stop its spread

However American Immigration Group is using the swine virus as an argument to close the US Mexico borders. see below

http://www.time.com/time/nation/article/0,8599,1894775,00.html

Ironically, Dr. Carlos del Rio, chairman of the global health department at Emory University has said that in case of serious outbreak of the swine flu, US would need mechanical ventilators which are made in Mexico

My first model - A Project for “Project Management & Collaboration” Elective

One of our deliverables for our “Project Management & Collaboration” elective was a project for Emory Health Care’s Strategic Planning Office. The project was to interview the team, and identify inefficient processes in their project management tasks and assist them in rectifying these issues. Our discussions with the team had led us to identify one of the problems that was “Constantly shifting priorities based on stakeholder positions lead to lack of consistency in project prioritization.”
We believed that the Strategic Planning Office needed our help in the area of project prioritization. Currently the SPO is managing 41 major projects that are scheduled for the next five months. These projects span across 3 different organizations. It was imperative for SPO to effectively effectively prioritize work efforts in order to finish key projects on time.
We noticed that project priorities were re-shuffled frequently , due to politics or system requirements. These reshuffles were made without evaluating how it would impact the project portfolio etc. We determined that a way to help the Strategic Management Office overcome this issue would be to develop a simple model that could quantitatively evaluate projects on multiple variables and generate a project prioritization ranking for each project in the portfolio.
With this in mind we created a model that allows users to prioritize projects based on attributes such as ROI, Alignment with Strategic Objective etc. For each attribute, users enter a number between 0 and 100, representing the variable’s utility. It is possible for some attributes, at a higher or a lower value, to have a diminishing marginal utility. Using utility instead of the actual value for a variable such as ROI will accommodate this diminishing marginal utility. Below is a screenshot of our model template. The model then applies a defined weighting to each attribute to calculate the overall score for each project, and then ranks the projects.

Impact of fear on Decision Making

An interesting article that talks about how Fear impact decision making. This is due to "endowment effect" which means that people tend to value objects a lot more than everyone else.

http://www.nytimes.com/2008/12/07/jobs/07pre.html.

Upon introspection this article makes sense regarding some of the decisions that I make when a lot is at stake. If we take fear out of the equation, decision making would be a lot efficient.

Decision Support Systems

http://www.jamaica-gleaner.com/gleaner/20090416/eyes/eyes2.html

An interesting article on combining different modules into one Integrated Decision Support System. Sounds very neat and also lists the possible challenges. The interesting aspect is that there is no mention for formulation of fundamental objectives. I believe for organizations, it just boils down to “increase share holder’s value”
. But does this - level the plain field for competitive organizations, or would the amount of data available would result in better optimization?

App to help you avoid make the “WRONG” decision

http://i.gizmodo.com/5195684/bad-decision-blocker-prevents-you-from-drunk-dialing-your-exes

Funny, but could be useful. I see this very similar to deletion checks on web pages that we put for users, to avoid an unwanted situation. By simplifying the scenario as a binomial distribution ( right and wrong decision). It would be interesting to calculate the change/reduction in probability of the wrong decision by using these tools/checks and the utility of wrong decision, which in case of the phone app could be anywhere from awkward silence to extreme embarrassment. In case of deletion of data, it could be anywhere from a simple slap on wrist by supervisor to job loss.

North Korea missile launch

http://businessmirror.com.ph/home/world/8395-us-outlines-plans-to-curb-nuke-weapons-.html

Recently North Korea launched a multistage nuclear capable missile. The missile launched failed and resulted in North Korea being in a worse situation in before because of 2 main reasons
- Sanctions against them
- Now everyone is aware that their missiles do not work, so the threat level is lowered.
They (N. Koreans) believed that the missile would be a good outcome because of the following 3 reasons
- They will be in a better position for negotiations with US and S. Korea
- global advertisement for its arms sales to developing countries.
- Generate public support

Their decision tree looked something like this -

With a successful missile launch results in 100, and a failure results in -100 and no missile launch maintains the status quo at 0. The issue was with the % of success of missile launch. I believe their % of probability should be greater than 0.5 since the first missile “Taepodong 1” was a failure.

Which way to drive to Emory

I usually use one of the following ways to drive to Emory -

- Freedom Pkwy (75S-> Freedom Pkwy->Mooreland->Briarcliff) (PATH A)
- Lavista (75S->85N-> Lavista->Clifton) (PATH B)
- N. Druid hills (75S->85N-> N. Druid Hills->Briarcliff) (PATH C)

Below is a model that I have created using the following 3 attributes
- Average time taken for travel
- % of a possible congestion
- Ease of driving

For average time I created a utility lookup table which is the following utility curve based on my preferences.



Similarly I created a utility lookup table for % of possible congestion see below



Ease of driving is made up of # of red lights(less are preferred), # of lanes of traffic (more are preferred),


After plotting the values of utilities for these 3 attributes I created the series graph and noticed that Path A is dominated by both Path B and Path C.




So I should choose between path A and path B. If I value .1 utility of Average time i.e. decrease in Average time from 42 minutes to 40 minutes same as decrease in .6 drop in utility i.e. increase of % possibility of congestion from 6% to 10% then I should be indifferent between these paths.
~ (0.1)x = (0.6)y
--> x = 6y
Where x is the utility of average time and y is the utility of probability of congestion.

However if my valuation is not the same as x= 6y when then I could arrive at an optimum path.