Machine learning is a rapidly growing field, with no doubt in part due to its countless applications to other fields, including pedagogy and the creation of computer-aided tutoring systems. To extend the functionality of FACT, an automated teaching assistant, we want to predict, using metadata produced by student activity, whether a student is capable of fixing their own mistakes. Logs were collected from previous FACT trials with middle school math teachers and students. The data was converted to time series sequences for deep learning, and ordinary features were extracted for statistical machine learning. Ultimately, deep learning models attained an accuracy of 60%, while tree-based methods attained an accuracy of 65%, showing that some correlation, although small, exists between how a student fixes their mistakes and whether their correction is correct.

The process of learning a new skill can be time consuming and difficult for both the teacher and the student, especially when it comes to computer modeling. With so many terms and functionalities to familiarize oneself with, this task can be overwhelming to even the most knowledgeable student. The purpose of this paper is to describe the methodology used in the creation of a new set of curricula for those attempting to learn how to use the Dynamic Traffic Simulation Package with Multi-Resolution Modeling. The current DLSim curriculum currently relates information via high-concept terms and complicated graphics. The information in this paper aims to provide a streamlined set of curricula for new users of DLSim, including lesson plans and improved infographics.

The HLA, Human Leukocyte Antigens, are encoded by a polymorphic set of genes where even a single base change can impact the function of the body’s immune response to foreign antigens [1]. Although many methods exist to type these alleles using whole-genome sequencing (WGS), few can use RNA sequencing (RNA-seq) to show the functional expression of the alleles with its inconsistency in coverage, and none of these allow for novel allele discovery. We present an approach using partially ordered graphs to project sequenced data onto the known alleles allowing for accurate and efficient typing of the HLA genes with flexibility for discovering new alleles and tolerance for poor sequence quality. This graph-guided approach to assembling and typing the HLA genes from RNA-seq has applications throughout precision medicine, facilitating the prevention and treatment of autoimmune diseases where allele expression can change. It is also a necessary step for determining donors for organ transplants with the least likelihood of rejection. This novel approach of combining database matching with partially ordered graphs for assembling genetic sequences of RNA-seq data could be applied towards typing other alleles.

Recent advancements in machine learning methods have allowed companies to develop advanced computer vision aided production lines that take advantage of the raw and labeled data captured by high-definition cameras mounted at vantage points in their factory floor. We experiment with two different methods of developing one such system to automatically track key components on a production line. By tracking the state of these key components using object detection we can accurately determine and report production line metrics like part arrival and start/stop times for key factory processes. We began by collecting and labeling raw image data from the cameras overlooking the factory floor. Using that data we trained two dedicated object detection models. Our training utilized transfer learning to start from a Faster R-CNN ResNet model trained on Microsoft’s COCO dataset. The first model we developed is a binary classifier that detects the state of a single object while the second model is a multiclass classifier that detects the state of two distinct objects on the factory floor. Both models achieved over 95% classification and localization accuracy on our test datasets. Having two additional classes did not affect the classification or localization accuracy of the multiclass model compared to the binary model.