A Taxonomy of Property Measures to Unify Active Learning and Human-centered Approaches to Data Labeling

2021 ◽  
Vol 11 (3-4) ◽  
pp. 1-42
Author(s):  
Jürgen Bernard ◽  
Marco Hutter ◽  
Michael Sedlmair ◽  
Matthias Zeppelzauer ◽  
Tamara Munzner
Keyword(s):  
Machine Learning ◽  
Active Learning ◽  
Visual Analytics ◽  
Building Blocks ◽  
Instance Selection ◽  
Four Dimensions ◽  
Fine Grained ◽  
Selection Strategies ◽  
Show Evidence ◽  
Future Work

Strategies for selecting the next data instance to label, in service of generating labeled data for machine learning, have been considered separately in the machine learning literature on active learning and in the visual analytics literature on human-centered approaches. We propose a unified design space for instance selection strategies to support detailed and fine-grained analysis covering both of these perspectives. We identify a concise set of 15 properties, namely measureable characteristics of datasets or of machine learning models applied to them, that cover most of the strategies in these literatures. To quantify these properties, we introduce Property Measures (PM) as fine-grained building blocks that can be used to formalize instance selection strategies. In addition, we present a taxonomy of PMs to support the description, evaluation, and generation of PMs across four dimensions: machine learning (ML) Model Output , Instance Relations , Measure Functionality , and Measure Valence . We also create computational infrastructure to support qualitative visual data analysis: a visual analytics explainer for PMs built around an implementation of PMs using cascades of eight atomic functions. It supports eight analysis tasks, covering the analysis of datasets and ML models using visual comparison within and between PMs and groups of PMs, and over time during the interactive labeling process. We iteratively refined the PM taxonomy, the explainer, and the task abstraction in parallel with each other during a two-year formative process, and show evidence of their utility through a summative evaluation with the same infrastructure. This research builds a formal baseline for the better understanding of the commonalities and differences of instance selection strategies, which can serve as the stepping stone for the synthesis of novel strategies in future work.

Machine Learning Approaches for Bangla Statistical Machine Translation

2013 ◽  
pp. 79-95
Author(s):  
Maxim Roy
Keyword(s):  
Machine Learning ◽  
Active Learning ◽  
Machine Translation ◽  
Language Processing ◽  
Statistical Machine Translation ◽  
Low Density ◽  
Learning Approaches ◽  
Translation Quality ◽  
Selection Strategies ◽  
Translation Accuracy

Machine Translation (MT) from Bangla to English has recently become a priority task for the Bangla Natural Language Processing (NLP) community. Statistical Machine Translation (SMT) systems require a significant amount of bilingual data between language pairs to achieve significant translation accuracy. However, being a low-density language, such resources are not available in Bangla. In this chapter, the authors discuss how machine learning approaches can help to improve translation quality within as SMT system without requiring a huge increase in resources. They provide a novel semi-supervised learning and active learning framework for SMT, which utilizes both labeled and unlabeled data. The authors discuss sentence selection strategies in detail and perform detailed experimental evaluations on the sentence selection methods. In semi-supervised settings, reversed model approach outperformed all other approaches for Bangla-English SMT, and in active learning setting, geometric 4-gram and geometric phrase sentence selection strategies proved most useful based on BLEU score results over baseline approaches. Overall, in this chapter, the authors demonstrate that for low-density language like Bangla, these machine-learning approaches can improve translation quality.

scholarly journals Interfaces for Searching and Triaging Large Document Sets: An Ontology-Supported Visual Analytics Approach

Information ◽  
2021 ◽  
Vol 13 (1) ◽  
pp. 8
Author(s):  
Jonathan Demelo ◽  
Kamran Sedig
Keyword(s):  
Machine Learning ◽  
Information Search ◽  
Information Seeking ◽  
Visual Analytics ◽  
Plug And Play ◽  
Multi Stage ◽  
Formative Evaluations ◽  
Future Work ◽  
Domain Independent ◽  
Visual Interface

We investigate the design of ontology-supported, progressively disclosed visual analytics interfaces for searching and triaging large document sets. The goal is to distill a set of criteria that can help guide the design of such systems. We begin with a background of information search, triage, machine learning, and ontologies. We review research on the multi-stage information-seeking process to distill the criteria. To demonstrate their utility, we apply the criteria to the design of a prototype visual analytics interface: VisualQUEST (Visual interface for QUEry, Search, and Triage). VisualQUEST allows users to plug-and-play document sets and expert-defined ontology files within a domain-independent environment for multi-stage information search and triage tasks. We describe VisualQUEST through a functional workflow and culminate with a discussion of ongoing formative evaluations, limitations, future work, and summary.

Efficient Prediction of Structural and Electronic Properties of Hybrid 2D Materials Using DFT and Machine Learning

2018 ◽  
Author(s):  
Sherif Tawfik ◽  
Olexandr Isayev ◽  
Catherine Stampfl ◽  
Joseph Shapter ◽  
David Winkler ◽  
...  
Keyword(s):  
Machine Learning ◽  
Band Gap ◽  
Density Functional ◽  
2D Materials ◽  
Van Der Waals ◽  
Building Blocks ◽  
Machine Learning Techniques ◽  
Interlayer Distance ◽  
Computational Screening ◽  
Wide Range

Materials constructed from different van der Waals two-dimensional (2D) heterostructures offer a wide range of benefits, but these systems have been little studied because of their experimental and computational complextiy, and because of the very large number of possible combinations of 2D building blocks. The simulation of the interface between two different 2D materials is computationally challenging due to the lattice mismatch problem, which sometimes necessitates the creation of very large simulation cells for performing density-functional theory (DFT) calculations. Here we use a combination of DFT, linear regression and machine learning techniques in order to rapidly determine the interlayer distance between two different 2D heterostructures that are stacked in a bilayer heterostructure, as well as the band gap of the bilayer. Our work provides an excellent proof of concept by quickly and accurately predicting a structural property (the interlayer distance) and an electronic property (the band gap) for a large number of hybrid 2D materials. This work paves the way for rapid computational screening of the vast parameter space of van der Waals heterostructures to identify new hybrid materials with useful and interesting properties.

Machine Learning Based Predictive Action on Categorical Non-Sequential Data

2020 ◽  
Vol 13 (5) ◽  
pp. 1020-1030
Author(s):  
Pradeep S. ◽  
Jagadish S. Kallimani
Keyword(s):  
Machine Learning ◽  
Data Analysis ◽  
Categorical Data ◽  
Numerical Data ◽  
Processing Technique ◽  
Machine Learning Algorithms ◽  
Sequential Data ◽  
Industry Standard ◽  
Robust Model ◽  
Future Work

Background: With the advent of data analysis and machine learning, there is a growing impetus of analyzing and generating models on historic data. The data comes in numerous forms and shapes with an abundance of challenges. The most sorted form of data for analysis is the numerical data. With the plethora of algorithms and tools it is quite manageable to deal with such data. Another form of data is of categorical nature, which is subdivided into, ordinal (order wise) and nominal (number wise). This data can be broadly classified as Sequential and Non-Sequential. Sequential data analysis is easier to preprocess using algorithms. Objective: The challenge of applying machine learning algorithms on categorical data of nonsequential nature is dealt in this paper. Methods: Upon implementing several data analysis algorithms on such data, we end up getting a biased result, which makes it impossible to generate a reliable predictive model. In this paper, we will address this problem by walking through a handful of techniques which during our research helped us in dealing with a large categorical data of non-sequential nature. In subsequent sections, we will discuss the possible implementable solutions and shortfalls of these techniques. Results: The methods are applied to sample datasets available in public domain and the results with respect to accuracy of classification are satisfactory. Conclusion: The best pre-processing technique we observed in our research is one hot encoding, which facilitates breaking down the categorical features into binary and feeding it into an Algorithm to predict the outcome. The example that we took is not abstract but it is a real – time production services dataset, which had many complex variations of categorical features. Our Future work includes creating a robust model on such data and deploying it into industry standard applications.

Alternatives

2020 ◽  
pp. 112-138
Author(s):  
Holly M. Smith
Keyword(s):  
Action Theory ◽  
Human Action ◽  
Fine Grained ◽  
Moral Properties ◽  
Future Work

Consequentialists have long debated (as deontologists should) how to define an agent’s alternatives, given that (a) at any particular time an agent performs numerous “versions” of actions, (b) an agent may perform several independent co-temporal actions, and (c) an agent may perform sequences of actions. We need a robust theory of human action to provide an account of alternatives that avoids previously debated problems. After outlining Alvin Goldman’s action theory (which takes a fine-grained approach to act individuation) and showing that the agent’s alternatives must remain invariant across different normative theories, I address issue (a) by arguing that an alternative for an agent at a time is an entire “act tree” performable by her, rather than any individual act token. I argue further that both tokens and trees must possess moral properties, and I suggest principles governing how these are inherited among trees and tokens. These proposals open a path for future work addressing issues (b) and (c).

scholarly journals How to measure uncertainty in uncertainty sampling for active learning

2021 ◽  
Author(s):  
Vu-Linh Nguyen ◽  
Mohammad Hossein Shaker ◽  
Eyke Hüllermeier
Keyword(s):  
Machine Learning ◽  
Active Learning ◽  
Sampling Strategies ◽  
Total Uncertainty ◽  
Uncertainty Sampling ◽  
Different Types ◽  
Alternative Approaches ◽  
Active Learner ◽  
Probabilistic Nature ◽  
Different Sources

AbstractVarious strategies for active learning have been proposed in the machine learning literature. In uncertainty sampling, which is among the most popular approaches, the active learner sequentially queries the label of those instances for which its current prediction is maximally uncertain. The predictions as well as the measures used to quantify the degree of uncertainty, such as entropy, are traditionally of a probabilistic nature. Yet, alternative approaches to capturing uncertainty in machine learning, alongside with corresponding uncertainty measures, have been proposed in recent years. In particular, some of these measures seek to distinguish different sources and to separate different types of uncertainty, such as the reducible (epistemic) and the irreducible (aleatoric) part of the total uncertainty in a prediction. The goal of this paper is to elaborate on the usefulness of such measures for uncertainty sampling, and to compare their performance in active learning. To this end, we instantiate uncertainty sampling with different measures, analyze the properties of the sampling strategies thus obtained, and compare them in an experimental study.

scholarly journals A Machine Vision Approach for Bioreactor Foam Sensing

2021 ◽  
pp. 247263032110088
Author(s):  
Jonas Austerjost ◽  
Robert Söldner ◽  
Christoffer Edlund ◽  
Johan Trygg ◽  
David Pollard ◽  
...  
Keyword(s):  
Machine Learning ◽  
Machine Vision ◽  
State Of The Art ◽  
Low Cost ◽  
High Accuracy ◽  
Consumer Electronics ◽  
Learning System ◽  
Automotive Applications ◽  
Fine Grained

Machine vision is a powerful technology that has become increasingly popular and accurate during the last decade due to rapid advances in the field of machine learning. The majority of machine vision applications are currently found in consumer electronics, automotive applications, and quality control, yet the potential for bioprocessing applications is tremendous. For instance, detecting and controlling foam emergence is important for all upstream bioprocesses, but the lack of robust foam sensing often leads to batch failures from foam-outs or overaddition of antifoam agents. Here, we report a new low-cost, flexible, and reliable foam sensor concept for bioreactor applications. The concept applies convolutional neural networks (CNNs), a state-of-the-art machine learning system for image processing. The implemented method shows high accuracy for both binary foam detection (foam/no foam) and fine-grained classification of foam levels.

scholarly journals Early Detection of Exposure to Toxic Chemicals Using Continuously Recorded Multi-Sensor Physiology

Sensors ◽  
2021 ◽  
Vol 21 (11) ◽  
pp. 3616
Author(s):  
Jan Ubbo van Baardewijk ◽  
Sarthak Agarwal ◽  
Alex S. Cornelissen ◽  
Marloes J. A. Joosen ◽  
Jiska Kentrop ◽  
...  
Keyword(s):  
Machine Learning ◽  
Early Detection ◽  
Physiological Signals ◽  
Machine Learning Techniques ◽  
Detection Accuracy ◽  
Chemical Agent ◽  
Counter Measures ◽  
Detection And Identification ◽  
Life Saving ◽  
Future Work

Early detection of exposure to a toxic chemical, e.g., in a military context, can be life-saving. We propose to use machine learning techniques and multiple continuously measured physiological signals to detect exposure, and to identify the chemical agent. Such detection and identification could be used to alert individuals to take appropriate medical counter measures in time. As a first step, we evaluated whether exposure to an opioid (fentanyl) or a nerve agent (VX) could be detected in freely moving guinea pigs using features from respiration, electrocardiography (ECG) and electroencephalography (EEG), where machine learning models were trained and tested on different sets (across subject classification). Results showed this to be possible with close to perfect accuracy, where respiratory features were most relevant. Exposure detection accuracy rose steeply to over 95% correct during the first five minutes after exposure. Additional models were trained to correctly classify an exposed state as being induced either by fentanyl or VX. This was possible with an accuracy of almost 95%, where EEG features proved to be most relevant. Exposure detection models that were trained on subsets of animals generalized to subsets of animals that were exposed to other dosages of different chemicals. While future work is required to validate the principle in other species and to assess the robustness of the approach under different, realistic circumstances, our results indicate that utilizing different continuously measured physiological signals for early detection and identification of toxic agents is promising.

Sign in / Sign up

Export Citation Format

Share Document