Electrical Engineering
      and Computer Sciences

Electrical Engineering and Computer Sciences

COLLEGE OF ENGINEERING

UC Berkeley

ML-o-scope: a diagnostic visualization system for deep machine learning pipelines

Daniel Bruckner

EECS Department
University of California, Berkeley
Technical Report No. UCB/EECS-2014-99
May 16, 2014

http://www.eecs.berkeley.edu/Pubs/TechRpts/2014/EECS-2014-99.pdf

The recent success of deep learning is driving a trend towards structurally complex computer vision models that combine feature extraction with predictive elements into integrated pipelines. While some of these models have achieved breakthrough results in applications like object recognition, they are difficult to design and tune, impeding progress. We feel that visual analysis can be a powerful tool to aid iterative development of deep model pipelines. Building on feature evaluation work in the computer vision community, we introduce ML-o-scope, an interactive visualization system for exploratory analysis of convolutional neural networks, a prominent type of pipelined model. We present ML-o-scope’s time-lapse engine that provides views into model dynamics during training, and evaluate the system as a support for tuning large scale object-classification pipelines.

Advisor: Michael Franklin


BibTeX citation:

@mastersthesis{Bruckner:EECS-2014-99,
    Author = {Bruckner, Daniel},
    Title = {ML-o-scope: a diagnostic visualization system for deep machine learning pipelines},
    School = {EECS Department, University of California, Berkeley},
    Year = {2014},
    Month = {May},
    URL = {http://www.eecs.berkeley.edu/Pubs/TechRpts/2014/EECS-2014-99.html},
    Number = {UCB/EECS-2014-99},
    Abstract = {The recent success of deep learning is driving a trend towards structurally complex computer vision models that combine feature extraction with predictive elements into integrated pipelines. While some of these models have achieved breakthrough results in applications like object recognition, they are difficult to design and tune, impeding progress. We feel that visual analysis can be a powerful tool to aid iterative development of deep model pipelines. Building on feature evaluation work in the computer vision community, we introduce ML-o-scope, an interactive visualization system for exploratory analysis of convolutional neural networks, a prominent type of pipelined model. We present ML-o-scope’s time-lapse engine that provides views into model dynamics during training, and evaluate the system as a support for tuning large scale object-classification pipelines.}
}

EndNote citation:

%0 Thesis
%A Bruckner, Daniel
%T ML-o-scope: a diagnostic visualization system for deep machine learning pipelines
%I EECS Department, University of California, Berkeley
%D 2014
%8 May 16
%@ UCB/EECS-2014-99
%U http://www.eecs.berkeley.edu/Pubs/TechRpts/2014/EECS-2014-99.html
%F Bruckner:EECS-2014-99