Forward

With the Q (and a few days of recovery) behind, I'll get back to regular posts. Here's the plan for this spring:

• MA5140 - Set Theory. This is pretty deep water. The class is a mix of Masters and PhD students and there's a significant gap between those two groups at UMSL. This material isn't particularly relevant to my thesis, but it's stuff I really like, so I plan on considering all the material, even if most of the class has to stay closer to shore.
• CS5320 - Evolutionary Computing. This could have been a pretty easy class had I not decided to include a substantial piece of my thesis work as the term project. As I did do that, I expect this to be a very time consuming class. Things that need to be developed:
• Update the data layer from CISS (my stratified query algorithm that may or may not get published) to do a better job of managing attributes in a block.
• Update the query engine from CISS to use the enhanced attribute information to determine if a block contains any rows in a query (and ignore it if it doesn't).
• Collect and store statistics on query patterns and block usage for each pattern.
• Build a test harness that can submit several thousand queries against the data using patterns that vary over time.
• Build a data ingestion program that sorts new data into blocks based on block preferences.
• Design an evolutionary algorithm that determines optimal block preferences.
• Implement the above algorithm and call it from the ingestion program.
• Build a secondary test harness that drives the block deletion/re-ingestion strategy between sessions of query harness.
• Write up the results and present to the class.
• Each of the above tasks is what we would call a 3-5 point story at work, meaning that a competent developer should be able to get it done in around a week. There are 9 stories in the list. There are 12 weeks remaining in the term. Oh, yeah, I have a day job, too. Seems I'll be busy. I may be able to cut the scope a bit and still have a good project and something useful going forward.
• Fortunately, I don't have any other school obligations for this semester. My adviser does seem to have renewed interest in getting CISS published, but that paper is pretty much ready to go. The term project for CS5320 will constitute adequate progress against the thesis.
• Also fortunate is that spring is the less nutty time at work, so I can take a few days off here and there if needed. I also don't much care how fast I run Boston, so running will simply be a fun thing to do without the stress of having to get in productive workouts.