Towards Human-Centered AI: Psychological concepts as foundation for empirical XAI research

2021 ◽  
Vol 0 (0) ◽  
Author(s):  
Katharina Weitz
Keyword(s):  
Machine Learning ◽  
Mental Models ◽  
Empirical Studies ◽  
End Users ◽  
User Needs ◽  
Specific Task ◽  
Explainable Ai ◽  
First Results ◽  
Shed Light ◽  
Psychological Concepts

Abstract Human-Centered AI is a widely requested goal for AI applications. To reach this is explainable AI promises to help humans to understand the inner workings and decisions of AI systems. While different XAI techniques have been developed to shed light on AI systems, it is still unclear how end-users with no experience in machine learning perceive these. Psychological concepts like trust, mental models, and self-efficacy can serve as instruments to evaluate XAI approaches in empirical studies with end-users. First results in applications for education, healthcare, and industry suggest that one XAI does not fit all. Instead, the design of XAI has to consider user needs, personal background, and the specific task of the AI system.

Understanding Machine Learning for Diversified Portfolio Construction by Explainable AI

2020 ◽  
Author(s):  
Markus Jaeger ◽  
Stephan Krügel ◽  
Dimitri Marinelli ◽  
Jochen Papenbrock ◽  
Peter Schwendner
Keyword(s):  
Machine Learning ◽  
Portfolio Construction ◽  
Explainable Ai ◽  
Diversified Portfolio

A Review on Machine-Learning Based Code Smell Detection Techniquesin Object-Oriented Software System(s)

2020 ◽  
Vol 13 ◽  
Author(s):  
Amandeep Kaur ◽  
Sushma Jain ◽  
Shivani Goel ◽  
Gaurav Dhiman
Keyword(s):  
Machine Learning ◽  
Programming Languages ◽  
Software Quality ◽  
Empirical Studies ◽  
Statistical Testing ◽  
Machine Learning Techniques ◽  
Support Vector ◽  
Code Smells ◽  
Detection Techniques ◽  
Code Smell

Context: Code smells are symptoms, that something may be wrong in software systems that can cause complications in maintaining software quality. In literature, there exists many code smells and their identification is far from trivial. Thus, several techniques have also been proposed to automate code smell detection in order to improve software quality. Objective: This paper presents an up-to-date review of simple and hybrid machine learning based code smell detection techniques and tools. Methods: We collected all the relevant research published in this field till 2020. We extracted the data from those articles and classified them into two major categories. In addition, we compared the selected studies based on several aspects like, code smells, machine learning techniques, datasets, programming languages used by datasets, dataset size, evaluation approach, and statistical testing. Results: Majority of empirical studies have proposed machine- learning based code smell detection tools. Support vector machine and decision tree algorithms are frequently used by the researchers. Along with this, a major proportion of research is conducted on Open Source Softwares (OSS) such as, Xerces, Gantt Project and ArgoUml. Furthermore, researchers paid more attention towards Feature Envy and Long Method code smells. Conclusion: We identified several areas of open research like, need of code smell detection techniques using hybrid approaches, need of validation employing industrial datasets, etc.

Explainable AI Models of Stock Crashes: A Machine-Learning Explanation of the Covid March 2020 Equity Meltdown

2021 ◽  
Author(s):  
Jean-Jacques Ohana ◽  
Steve Ohana ◽  
Eric Benhamou ◽  
David Saltiel ◽  
Beatrice Guez
Keyword(s):  
Machine Learning ◽  
Explainable Ai

scholarly journals Software Testing: Issues and Challenges of Artificial Intelligence & Machine Learning

2021 ◽  
Vol 12 (1) ◽  
pp. 101-112
Author(s):  
Kishore Sugali ◽  
Chris Sprunger ◽  
Venkata N Inukollu
Keyword(s):  
Artificial Intelligence ◽  
Machine Learning ◽  
Software Testing ◽  
Future Research ◽  
Critical Component ◽  
Development Methodology ◽  
Testing Strategies ◽  
History Of ◽  
Shed Light ◽  
The Way

The history of Artificial Intelligence and Machine Learning dates back to 1950’s. In recent years, there has been an increase in popularity for applications that implement AI and ML technology. As with traditional development, software testing is a critical component of an efficient AI/ML application. However, the approach to development methodology used in AI/ML varies significantly from traditional development. Owing to these variations, numerous software testing challenges occur. This paper aims to recognize and to explain some of the biggest challenges that software testers face in dealing with AI/ML applications. For future research, this study has key implications. Each of the challenges outlined in this paper is ideal for further investigation and has great potential to shed light on the way to more productive software testing strategies and methodologies that can be applied to AI/ML applications.

scholarly journals A user-friendly tool for cloud-based whole slide image segmentation, with examples from renal histopathology

2021 ◽  
Author(s):  
Brendon R Lutnick ◽  
David Manthey ◽  
Jan U Becker ◽  
Brandon Ginley ◽  
Katharina Moos ◽  
...  
Keyword(s):  
Machine Learning ◽  
User Interface ◽  
Interstitial Fibrosis ◽  
End Users ◽  
Great Promise ◽  
Histological Structure ◽  
Full Potential ◽  
Learning Tools ◽  
Tubular Atrophy ◽  
Slide Image

Image-based machine learning tools hold great promise for clinical applications in nephropathology and kidney research. However, the ideal end-users of these computational tools (e.g., pathologists and biological scientists) often face prohibitive challenges in using these tools to their full potential, including the lack of technical expertise, suboptimal user interface, and limited computation power. We have developed Histo-Cloud, a tool for segmentation of whole slide images (WSIs) that has an easy-to-use graphical user interface. This tool runs a state-of-the-art convolutional neural network (CNN) for segmentation of WSIs in the cloud and allows the extraction of features from segmented regions for further analysis. By segmenting glomeruli, interstitial fibrosis and tubular atrophy, and vascular structures from renal and non-renal WSIs, we demonstrate the scalability, best practices for transfer learning, and effects of dataset variability. Finally, we demonstrate an application for animal model research, analyzing glomerular features in murine models of aging, diabetic nephropathy, and HIV associated nephropathy. The ability to access this tool over the internet will facilitate widespread use by computational non-experts. Histo-Cloud is open source and adaptable for segmentation of any histological structure regardless of stain. Histo-Cloud will greatly accelerate and facilitate the generation of datasets for machine learning in the analysis of kidney histology, empowering computationally novice end-users to conduct deep feature analysis of tissue slides.

scholarly journals PyDL8.5: a Library for Learning Optimal Decision Trees

2020 ◽  
Author(s):  
Gaël Aglin ◽  
Siegfried Nijssen ◽  
Pierre Schaus
Keyword(s):  
Machine Learning ◽  
Decision Trees ◽  
Efficient Algorithm ◽  
Optimal Decision ◽  
Learning Tasks ◽  
Explainable Ai ◽  
Classification Tasks ◽  
Interpretable Models ◽  
Limited Depth

Decision Trees (DTs) are widely used Machine Learning (ML) models with a broad range of applications. The interest in these models has increased even further in the context of Explainable AI (XAI), as decision trees of limited depth are very interpretable models. However, traditional algorithms for learning DTs are heuristic in nature; they may produce trees that are of suboptimal quality under depth constraints. We introduce PyDL8.5, a Python library to infer depth-constrained Optimal Decision Trees (ODTs). PyDL8.5 provides an interface for DL8.5, an efficient algorithm for inferring depth-constrained ODTs. The library provides an easy-to-use scikit-learn compatible interface. It cannot only be used for classification tasks, but also for regression, clustering, and other tasks. We introduce an interface that allows users to easily implement these other learning tasks. We provide a number of examples of how to use this library.

scholarly journals You Are the Only Possible Oracle: Effective Test Selection for End Users of Interactive Machine Learning Systems

2014 ◽  
Vol 40 (3) ◽  
pp. 307-323 ◽  
Author(s):  
Alex Groce ◽  
Todd Kulesza ◽  
Chaoqiang Zhang ◽  
Shalini Shamasunder ◽  
Margaret Burnett ◽  
...  
Keyword(s):  
Machine Learning ◽  
End Users ◽  
Learning Systems ◽  
Test Selection ◽  
Interactive Machine Learning ◽  
Selection For

Glean

2021 ◽  
Vol 14 (6) ◽  
pp. 997-1005
Author(s):  
Sandeep Tata ◽  
Navneet Potti ◽  
James B. Wendt ◽  
Lauro Beltrão Costa ◽  
Marc Najork ◽  
...  
Keyword(s):  
Machine Learning ◽  
Data Management ◽  
Real World ◽  
Empirical Studies ◽  
Ground Truth ◽  
Training Data ◽  
Ground Truth Data ◽  
Document Type ◽  
Machine Learning Model ◽  
Structured Information

Extracting structured information from templatic documents is an important problem with the potential to automate many real-world business workflows such as payment, procurement, and payroll. The core challenge is that such documents can be laid out in virtually infinitely different ways. A good solution to this problem is one that generalizes well not only to known templates such as invoices from a known vendor, but also to unseen ones. We developed a system called Glean to tackle this problem. Given a target schema for a document type and some labeled documents of that type, Glean uses machine learning to automatically extract structured information from other documents of that type. In this paper, we describe the overall architecture of Glean, and discuss three key data management challenges : 1) managing the quality of ground truth data, 2) generating training data for the machine learning model using labeled documents, and 3) building tools that help a developer rapidly build and improve a model for a given document type. Through empirical studies on a real-world dataset, we show that these data management techniques allow us to train a model that is over 5 F1 points better than the exact same model architecture without the techniques we describe. We argue that for such information-extraction problems, designing abstractions that carefully manage the training data is at least as important as choosing a good model architecture.

Sign in / Sign up

Export Citation Format

Share Document