In this paper we describe a system to augment human project managers of large scale software projects by predicting future SLA violations, using a machine learning approach. We detail the way we collected, processed, and formatted the data from a much larger database ill-suited for the problem at hand. We then detail how we extracted training examples through prediction engineering, as well as our semi-automated feature engineering approach. Our system was able to achieve a false positive rate of .311, compared to a non-machine learning baseline of and a simple machine learning baseline without feature engineering scoring .689 and .723 respectively...

In this paper, we propose a paradigm shift from the current practice of creating machine learning models that requires months-long discovery, exploration and “feasibility report” generation, followed by re-engineering for deployment, in favor of a rapid 8 week long process of development, understanding, validation and deployment that can executed by developers or subject matter experts (non-ML experts) using reusable APIs. It accomplishes what we call a “minimum viable data-driven model,” delivering a ready-to-use machine learning model for problems that haven’t been solved before using machine learning…

In this paper, we designed a formal language, called Trane, for describing prediction problems over relational datasets, implemented a system that allows data scientists to specify problems in that language. We show that this language is able to describe several prediction problems and even the ones on Kaggle- a data science competition website. We express 29 different Kaggle problems in this language. We designed an interpreter, which translates input from the user, specified in this language, into a series of transformation and aggregation operations…

In this paper, we present the concept of data science foundry for data from Massive Open Online Courses. In the foundry we present a series of software modules that transform the data into different representations. Ultimately, each online learner is represented using a set of variables that capture his/her online behavior. These variables are captured longitudinally over an interval. Using this representation we then build a predictive analytics stack that is able to predict online learners behavior as the course progresses in real time. To demonstrate the efficacy of the foundry, we attempt to solve an important prediction problem...

This paper outlines the design of a software system that enhances the workflow eficiency and usability of a liquid- handling robot for executing wet lab protocols, with a focus on synthetic biology. The system we propose builds on the BioCAD software suite...