Round Table: Dr. Amit Pande – Improving Mobile Applications Using Data-Driven Models
Improving Mobile Applications Using Data-Driven Models
Dr. Amit Pande
Research Scientist in Computer Science
University of California Davis
Cookies and coffee provided at 1:30 p.m. Talk begins at 2 p.m.
The vast amounts of data sensed and consumed in mobile devices/ sensors catalyze the need for developing methods to make it actionable as well as efficient and secure. The word ‘efficient’ implies reduced latency and resource-efficient implementation on mobile devices and the cloud. The heterogeneity of sensed data (video, network packets, sensor readings, text) and the cost of sensing, securing and transmitting the sensed data put additional constraints in developing a practical system using big-data analytics in the cloud. In my research I try to address these challenges. In this talk, I will talk about our two ongoing works in this domain.
In first work, we present a visual acuity framework which makes small online computations in mobile devices to get input features and gives an accurate estimate of video Quality of Experience (QoE) using Bagged Regression Tree. The prediction model has over 78% accuracy. Next, we envision collecting large volume of actual video watch data using a customized app and identify psycho-visual and other factors affecting video QoE. In other work, we try to study the Energy Expenditure (EE) and personal activity at individual and population level. We developed a model for EE estimation from mobile sensor readings using Bagged Regression Tree for diseased and healthy population. It yields up to 96% correlation with actual Energy Expenditure (EE).
Amit Pande is a research scientist in Department of Computer Science at the University of California Davis working with Prof. Prasant Mohapatra. Prior to that, he completed his PhD from Iowa State University (2010) in Computer Engineering with Dr. Joseph Zambreno and finished his Bachelors in Electronics and Communications Engineering from IIT Roorkee, India (2007). He has been recipient of NSF Computational Innovation Fellowship (2010-12), several Best Paper and Design Contest Awards as well as Research Excellence Awards from UC Davis, Iowa State University and IIT Roorkee. His research interests in data science applications, health informatics, mobile security and hardware acceleration.
Dr. Pande is a candidate for the Big Data faculty position.