Active Learning to Maximize Area Under the ROC Curve

In active learning, a machine learning algorithmis given an unlabeled set of examples U, and is allowed to request labels for a relatively small subset of U to use for training. The goal is then to judiciously choose which examples in U to have labeled in order to optimize some performance criterion, e.g. classification accuracy. We study how active learning affects AUC. We examine two existing algorithms from the literature and present our own active learning algorithms designed to maximize the AUC of the hypothesis. One of our algorithms was consistently the top performer, and Closest Sampling from the literature often came in second behind it. When good posterior probability estimates were available, our heuristics were by far the best.

Using Math Vocabulary Building to Increase Problem Solving Abilities in a 5th Grade Classroom

In this action research study of my 5th grade mathematics class, I investigated how students’ understanding of math vocabulary impacts their understanding of the curriculum. I discovered math vocabulary plays an important role in a student’s ability to understand daily lessons, complete homework, discuss ideas in groups, take tests and be successful on achievement tests. A student’s ability to understand the words around him (or her) in math class seem very related to his or her ability to solve word problems. Word problems are what our national assessments are all about. I also discovered that direct instruction and support of math vocabulary increased test scores and confidence in students as test takers. As a result of this research, I plan to continue to find ways to emphasize the vocabulary used in our current math curriculum. This process will start at the beginning of the year. I will continue to look for strategies that promote math vocabulary retention in my students. And finally, I will share my findings with my colleagues, so my research can be used as part of our School Improvement Goals.