cv | papers | pre-prints | books | tools | data
Director, RAISE lab, "Real-world AI for SE":
funding | projects | students
Here's my detailed cv.
For info on what I actually try to do and what I've achieved, then read below.
Here is a sample of my "less, but better" results (and for more, see Google Scholar):
2017Intelligent defect predictors can stop developers wasting time while they are fixing bad smells.
2015A simple feature and instance selectors let software projects share privatized data, without missing important patterns.
2012Very simple social metrics can generate near-optimal predictors for software quality.
2010Static code defect predictors have inherent limitations. But these limits can fixed via a new learner, very simple learner, that better understand the business goals.
2007Surprisingly effective defect predictors can be built from simple static code attributes.
2017Very simple optimizers can out perform overly complex deep learners.
2016Very simple optimizers can dramatically improve the performance of data miners learning software quality predictors.
2013Active learners can simplify and reduce the cost of search-based SE by orders of magnitude.
2002Search-based SE methods can easily and readily and critically assess long held SE truisms.
2017Even when project data collects data using different labels, we can still transfer lessons learned between them.
2016Ultra-simple transfer learning methods (called "bellwethers") enable effective transfer of lessons learned.
2013How to transfer lessons learned from past projects? Easy! Clustering tools enable transferring lessons learned between software projects.
2009A simple nearest neighbor relevancy filtering resulted in one of the first general results in software analytics: defect predictor learned from Turkish toasters could be successfully applied to NASA flight software (and vice versa).
2017Seemingly complex, conflicting models can be tamed and controlled via some simple stochastic probing.
2003Contrast set learners can explain enormous decision trees (6000 node) learned from complex requirements models just 6 rules.
2002Contrast set learners find simple controllers in requirements models.
2016The effort to build complex software can be estimated by very simple equations.
2013Active learners can easily estimate large software projects after just a few samples.
2016Text miners can succinctly summarize thousands of technical papers about SE.
2015Data miners can greatly simplify and reduce the effort involved in data collection for community health studies.
2010Simple contrast-set learners out-perform state-of-the-art optimizers for spacecraft control;
1990The lesson of decades of expert systems research is that, for specific domains, human expertise can be readily captured in just a few rules.