Feeds:
Posts
Comments

Archive for the ‘Analytics-ORMS’ Category

Since I am of that age… you know… when every friend is either getting married or having kids, when – at the sight of  you holding someone elses baby – someone is bound to say your are ‘rehearsing’ , when everyone is asking when are we expanding the family… I thought it would be appropriate to put in my 5 cents for the INFORMS Blog Challenge.

A few years ago, at the 2007 INFORMS Annual meeting in Seattle, I attended a keynote by Dr. Ralph L. Keeney centered  on using O.R. to improve a our quality of life.  Specifically, he talked  about a model that he and Professor Dinah A. Vernik had been working on called “Analysis of the Biological Clock Decision”.  This model examines when a woman should begin trying to conceive. Dr. Keeney presented a simplified spreadsheet model where users can set up most parameters related to a woman’s professional, social, and family aspects of life and integrates them into a quality-of-life curve that includes the changing relative importance of these aspects with age over a woman’s life. They also included the probabilities of conceiving when trying, as a function of a woman’s age.

Now, how cool is this?  This type of research shows us that operations research can be included in very intimate family decisions. However we must keep in mind that, like most mathematical models,  the model does not provide the one and holy answer to the question that haunts us, but instead it provides a framework to better answer the question, in this case a “formalized way to sort through conflicting pressures and considerations related to beginning a family”.

After going to this talk in Seattle I remember thinking that I would get home, find the spreadsheet in the link they provided and try to build a curve for myself. Of course back then, while I was still in grad school, the whole idea of kids hadn’t set seed in my mind yet so I never did look at the spreadsheet. It feels like it might be good idea to take a look in the next couple of years.

O.R. and my family

Since we are on the subject of family and O.R. I will share that my chosen family, meaning the one my husband and I started, only exists because of O.R. We were in fact, officially introduced by my former Operations Management professor, who happens to be a former INFORMS president. When we met I was a teaching assistant for the Optimization course and a marker for the Stochastic Modeling course and had already decided that O. R. was my thing.  My husband had already completed his MSc. in O. R., was working with my O. M. professor as a researcher and a lecturer and had already decided that he was eventually going to pursue a PhD. in O.R. which he is now soon to finish. During the summer we met I was to work on a O.R. related summer internship under my husband’s supervision. The rest of the story is very generic and not very O.R. related but from time to time I like to remember that O.R. is the reason we met.

Being in the same field means that we really get each others little work problems, like the bug on the code that took a whole day to find (ugh!). Also, the rest of our industrial engineer friends usually joke about how our household is likely to have a dynamic programing model to do the grocery shopping schedule with a policy for stock replenishment. They also joke about our kids are going to be born geeks and optimizers which would be fine with us.

If this O.R. family of two isn’t enough I will add that my brother-in-law also has a Ph.D in O.R.. He and his wife are also Industrial Engineers and so is my sister. My parents somewhat get what it is we do, explaining it wasn’t very hard since my dad used to be a traveling salesmen. You can imagine that finding an example to use was quite easy.

Learn more about the Analysis of the Biological Clock Decision:

Advertisements

Read Full Post »

What is the first thought that comes to your mind when you hear the words “Social Network”? I think of Facebook (due to the movie title perhaps), I also think of Twitter, LinkedIn, Google+, etc. I think of them even though I know these online social networking services are only applications of a much wider concept which considers any social structure made of individuals which are connected to each other. Social networks also exist in social settings such us our work, our group of friends and even within the members of our gym.

Historically we have never had thorough records of our own -or any- social networks. Although the gym is likely to have a record of all its members, the connections between members, such as common classes and preferred machines, are not necessarily recorded and therefore cannot be used to make decisions. We don’t even have data on far more interesting and analysis-worthy social networks like the ones describing the intricate relations between members of drug cartels or terrorist groups.

What makes Facebook et al. so neat is that we, the users, in our effort to exist and connect on cyberspace, create a huge amount of connectivity data that can be visualized and measured. Imagine this: every time you set up a profile you create a node that represents yourself, after that everything we do can be translated into links between nodes. Direct links are made when we connect to people (friends/circles/followers) but also indirect links are made through the things we like/+1/retweet, the places connect from and the words we say.

The business models behind these massive social networking sites rely on math models that analyze and use the data provided by each user. These models are able to place us into groups by looking at data entries such as our location, our ‘likes’ and even, through text analytics, our posts. They are able to measure how close or far away we are from other nodes in the network and are capable of telling us who we should connect to, decide which ads are of interest to us and show us which job posting we should apply for. It is clear that these major networking sites have found a way to use and profit from analytics.

Now let me tell you about my favorite social networking site: Ravelry (sorry Twitter/Facebook people!). Ravelry is an online community for yarn lovers. Users set up a profile, connect with friends and are able to upload info on every project they have worked on or are planning to start, on every needle they own and on every yarn skein in their stash. Additionally, within every project  users can rate the pattern and yarn used, enter information such start and end date, size made and the level of happiness with the result of the project. All of these entries create very rich database of projects and patterns, which is available to all users and can be explored through a search engine with thorough filtering capabilities. The beauty of this social network is that users are connected to each other not only through their friends and groups, but also through the yarns and patterns they work. Also, given the visibility of the data, crafters can use this social network to make decisions like what projects to start, what yarns to use and what modifications to make, all based on other peoples entries.

As cool as Revelry already is, I believe smaller and more focused social networking sites like this one, could benefit from advance Analytics, especially considering that a lot of the rich information that is being collected could be further analyzed and better displayed to deliver even more value to users. The very visible and vast amount of data available on this social network allows me to dream that someday we will have more numbers and analytics surrounding the world of crafts.

Read a great article on Ravelry by Slate here. Also, if you have an account, befriend me on Ravelry, I am MarianelaP

NOTE: I apologize for the lack of editing on this post. I was trying to write fast enough to be able submit this post for the July INFORMS Blog challenge. I promise to put more work into future posts.

Read Full Post »

%d bloggers like this: