Microsoft Technology Licensing, LLC

Yijin Wei

Neelakantan Sundaresan

Roshanak Zilouchian Moghaddam

Kalpathy Sitaraman Sivaraman

Shaun Miller

Arun Ramanathan Chandrasekhar

Claude Remillard

Mark Steven Schnitzer

Balan Subramanian

Yiping Dou

Tanmayee Prakash Kamath

Runtime errors in a source code program are detected in advance of execution by machine learning models. Features representing a context of a runtime error are extracted from source code programs to train a machine learning model, such as a random forest classifier, to predict the likelihood that a code snippet has a particular type of runtime error. The features are extracted from a syntax-type tree representation of each method in a program. A model is generated for distinct runtime errors, such as arithmetic overflow, and conditionally uninitialized variables.

Detection of runtime errors using machine learning

A cloud resource management system trains, through ensemble learning, multiple time series forecasting models to forecast a future idle time of a virtual machine operating on a cloud computing service. The models are trained on historical usage and metric data of the virtual machine. The metric data includes CPU usage, disk usage and network usage. A select one of the models having the best accuracy for a target virtual machine is used in a production run to predict when the virtual machine will be idle. At this time, the virtual machine may be automatically shutdown in order to reduce the expense associated with the continued operation of the virtual machine.

Cloud resource management using machine learning

Growing community of inventors

Bellevue, WA, United States of America

Yijin Wei