Title

Simplifying User Interfaces for Data Science and Machine Learning Applications

Presentation Type

Poster

Student

Yes

Track

Tools

Abstract

Machine learning is in a growing state as more businesses and individuals realize the power it carries in bringing a deeper understanding to large sums of existing data and make predictions based on discovered correlations that weren’t apparent before building a model. To accelerate the growth of this field, simplifying the process of machine learning will potentially lead to increasing the efficiency of the process of machine learning, in addition to lowering the ceiling of previous knowledge needed to start building models which will bring in more newcomers to the field of data science helping it grow as a community and a science. My research focuses on finding the implications of developing an open-source GUI (Graphical User Interface) set on top of a popular machine learning framework like TensorFlow.

Having a modular open-source GUI based machine learning system built to translate function calls to simple drag and drop operations that could be fit to use any of the popular machine learning specific and general data processing python modules could potentially accelerate the process of building models and reduce the number of human errors involved in manually writing python code. This approach to machine learning will also eliminate the need to learn to code for most of its applications, which in turn could bring in many new students who initially strayed away from the field due to having the requirement of having knowledge of coding concepts. Further research is needed to evaluate the time and cost needed to develop such a framework.

Start Date

2-11-2020 1:00 PM

This document is currently not available here.

Share

COinS
 
Feb 11th, 1:00 PM

Simplifying User Interfaces for Data Science and Machine Learning Applications

Volstorff A

Machine learning is in a growing state as more businesses and individuals realize the power it carries in bringing a deeper understanding to large sums of existing data and make predictions based on discovered correlations that weren’t apparent before building a model. To accelerate the growth of this field, simplifying the process of machine learning will potentially lead to increasing the efficiency of the process of machine learning, in addition to lowering the ceiling of previous knowledge needed to start building models which will bring in more newcomers to the field of data science helping it grow as a community and a science. My research focuses on finding the implications of developing an open-source GUI (Graphical User Interface) set on top of a popular machine learning framework like TensorFlow.

Having a modular open-source GUI based machine learning system built to translate function calls to simple drag and drop operations that could be fit to use any of the popular machine learning specific and general data processing python modules could potentially accelerate the process of building models and reduce the number of human errors involved in manually writing python code. This approach to machine learning will also eliminate the need to learn to code for most of its applications, which in turn could bring in many new students who initially strayed away from the field due to having the requirement of having knowledge of coding concepts. Further research is needed to evaluate the time and cost needed to develop such a framework.