Interactive visual machine learning in spreadsheets

@article{Sarkar2015InteractiveVM,
  title={Interactive visual machine learning in spreadsheets},
  author={Advait Sarkar and Mateja Jamnik and Alan F. Blackwell and Martin Spott},
  journal={2015 IEEE Symposium on Visual Languages and Human-Centric Computing (VL/HCC)},
  year={2015},
  pages={159-163}
}
  • Advait Sarkar, M. Jamnik, +1 author M. Spott
  • Published 17 December 2015
  • Computer Science
  • 2015 IEEE Symposium on Visual Languages and Human-Centric Computing (VL/HCC)
BrainCel is an interactive visual system for performing general-purpose machine learning in spreadsheets, building on end-user programming and interactive machine learning. BrainCel features multiple coordinated views of the model being built, explaining its current confidence in predictions as well as its coverage of the input domain, thus helping the user to evolve the model and select training examples. Through a study investigating users' learning barriers while building models using… Expand
A Review of User Interface Design for Interactive Machine Learning
TLDR
A structural and behavioural model of a generalised IML system is proposed and a solution principles for building effective interfaces for IML are identified, identified strands of user interface research key to unlocking more efficient and productive non-expert interactive machine learning applications. Expand
Constructivist Design for Interactive Machine Learning
TLDR
It is argued that the objectives of interactive machine learning can be interpreted as constructivist, and it is shown how constructivist learning environments pose critical questions for the design of interactiveMachine learning systems. Expand
Simpler Machine Learning Using Spreadsheets: Neural Network Predict
TLDR
This paper renders Neural Network, a key machine learning methodology, on the ubiquitous and easily comprehensible spreadsheet without macros or add-ins, employing only elementary operations and if so desired, optionally leveraging on its built-in Solver. Expand
Human-Machine Collaboration for Democratizing Data Science
TLDR
A novel framework and system that wants to democratize data science by allowing users to interact with standard spreadsheet software in order to perform and automate various data analysis tasks ranging from data wrangling, data selection, clustering, constraint learning, predictive modeling and auto-completion. Expand
Spreadsheet interfaces for usable machine learning
  • Advait Sarkar
  • Computer Science
  • 2015 IEEE Symposium on Visual Languages and Human-Centric Computing (VL/HCC)
  • 2015
TLDR
A line of research is presented into using the spreadsheet - already familiar to end-users as a paradigm for data manipulation - as a usable interface which lowers the statistical and computing knowledge barriers to building and using these models. Expand
On Building Design Guidelines for An Interactive Machine Learning Sandbox Application
TLDR
A sandbox approach with the goal of designing alternative programming interactions for machine learning tasks is introduced and a set of guidelines have been drafted and supported with user interviews to validate the proposed design framework. Expand
User Modelling for Avoiding Overfitting in Interactive Knowledge Elicitation for Prediction
TLDR
It is shown that the key idea of inferring user knowledge with probabilistic user models has general applicability in guarding against overfitting and improving interactive machine learning. Expand
VisualSynth: Democratizing Data Science in Spreadsheets
TLDR
This work introduces VisualSynth, a framework that wants to democratize data science by enabling naive end-users to specify the data science tasks that match their needs by highlighting parts of the data using colors. Expand
Children as data scientists : explorations in creating, thinking, and learning with data
TLDR
This thesis finds that children use Scratch Community Blocks not only to create with data through stories and games, but also to think with data by engaging in self-reflection about their own learning and social participation, and through critical conversations about the role of data within the culture of the Scratch community. Expand
Calculation View: multiple-representation editing in spreadsheets
TLDR
Through a user study, it is demonstrated that the use of multiple representations can significantly improve user performance when performing spreadsheet authoring and debugging tasks. Expand
...
1
2
...

References

SHOWING 1-10 OF 27 REFERENCES
Effective End-User Interaction with Machine Learning
TLDR
This work presents three explorations in designing for effective end-user interaction with machine learning in CueFlik, a system developed to support Web image search, demonstrating that interactions designed to balance the needs of end-users and machine learning algorithms can significantly improve the effectiveness ofend-user interactive machine learning. Expand
Principles of Explanatory Debugging to Personalize Interactive Machine Learning
TLDR
An empirical evaluation shows that Explanatory Debugging increased participants' understanding of the learning system by 52% and allowed participants to correct its mistakes up to twice as efficiently as participants using a traditional learning system. Expand
Designing for effective end-user interaction with machine learning
TLDR
This dissertation work aims to advance the understanding of how to design effective end-user interaction with interactive machine learning systems by investigating new techniques that move beyond naïve or ad-hoc approaches and balance the needs of both end-users and machine learning algorithms. Expand
Dis-function: Learning distance functions interactively
TLDR
It is illustrated empirically that with only a few iterations of interaction and optimization, a user can achieve a scatterplot view and its corresponding distance function that reflect the user's knowledge of the data. Expand
Teach and try: A simple interaction technique for exploratory data modelling by end users
TLDR
A spreadsheet-based interaction technique that can be used to build and apply sophisticated statistical models such as neural networks, decision trees, support vector machines and linear regression is described. Expand
Spreadsheet interfaces for usable machine learning
  • Advait Sarkar
  • Computer Science
  • 2015 IEEE Symposium on Visual Languages and Human-Centric Computing (VL/HCC)
  • 2015
TLDR
A line of research is presented into using the spreadsheet - already familiar to end-users as a paradigm for data manipulation - as a usable interface which lowers the statistical and computing knowledge barriers to building and using these models. Expand
Feedback-driven interactive exploration of large multidimensional data supported by visual classifier
TLDR
This work introduces a framework for a feedback-driven view exploration, inspired by relevance feedback approaches used in Information Retrieval, and presents an instantiation of the framework for exploration of Scatter Plot Spaces based on visual features. Expand
Structured labeling for facilitating concept evolution in machine learning
TLDR
This paper introduces the notion of concept evolution, the changing nature of a person's underlying concept which can result in inconsistent labels and thus be detrimental to machine learning, and introduces two structured labeling solutions. Expand
Scikit-learn: Machine Learning in Python
Scikit-learn is a Python module integrating a wide range of state-of-the-art machine learning algorithms for medium-scale supervised and unsupervised problems. This package focuses on bringingExpand
Confidence, command, complexity: metamodels for structured interaction with machine intelligence
TLDR
A speculative discussion of a potential solution: metamodels of machine cognition, where the notion of "correctness" for these programs is now unknown or ill-defined. Expand
...
1
2
3
...