The Edge Trends Committee sponsored a free seminar entitled “Risk Management Best Practices – What Would You Do?” on Thursday, April 23, 2015 at 7:30 a.m. in the UCSD Extension Classroom. The guest speaker was Dr. Pedro Maria-Sanchez, a certified Professional Risk Manager (PRM), and Professional Project Manager (PMP) with a Ph.D. in Risk Management. The 24 attendees were invited to submit questions for him in advance, which also entered them in a prize drawing. LAMP graduate Alex Corvino won a Fitbit Charge Wireless Activity Wristband!
LAMP alum Sunhyoung Han, Sr. Principal Scientist at ID Analytics, Inc. wrote a report on the seminar:
The seminar on Risk Management was interesting not because it is new concept but because it gave new perspective on how you define, recognize, and handle risks. Risk is associated with uncertain events with probability, and when it is handled with proper management it can be an opportunity rather than pure disaster. It is also interesting to see that people tend to be optimistic in planning phase, so even though risks are real and it happens all the time, not many people schedule considering unexpected events.
The first step for risk management is risk identification, and it is very important to have the risk statement in the beginning as specific as possible. Of course there will be updates as one gathers more data later, but trying to be specific with all the knowledge up to that point is crucial.
A well documented risk management plan not only reduces possibility for finger pointing when a problem happens, but also creates great way of communication. Risk management plans where one can see probability along with related costs in one site can easily lead to an optimal strategic plan, corresponding all possible scenario in the future. It is much more convincing to ask for resources when one has analysis chart to show “what if” cases attached with associated dollar loss.
Then the seminar made me think about how my company is doing regarding the risk management. We don’t have risk management planning, but we try to minimize planning span so that we can schedule with minimum uncertainty. One more solution to deal with risks: we have brief daily standup meeting to see if there is any blocker or something unexpected that happened the day before. In this way, we can adjust the plan on daily basis.
To take it one step further to risk management, there was the question “Can we have some predictive model to predict risk in the future?”, which originated from one of my fellow LAMPers. This was very interesting to hear, given that my job is data scientist who makes predictive model for various problems.
Machine Learning, also known as Artificial Intelligence, can be used as part of quantifying the problem. If it is trained properly it can be generate more precise probabilistic judgment on certain events than human expert can do. The challenge to applying the Machine Learning method to risk management is that we need to first quantify what experts know. Of course, the number of data points to train the algorithm is another issue, but quantifying what experts know as known as feature engineering is more critical part to solve.
Speak Your Mind