scholarly journals Machine Learning Based Overbound Modeling of Multipath Error for Safety Critical Urban Environment

2021 ◽  
Author(s):  
Heekwon No ◽  
Carl Milner
Keyword(s):  
Machine Learning ◽  
Urban Environment ◽  
Multipath Error ◽  
Safety Critical

scholarly journals Urban living environment assessment index system based on psychological security

2019 ◽  
Author(s):  
Xia Huiyi ◽  
◽  
Nankai Xia ◽  
Liu Liu ◽  
◽  
...  
Keyword(s):  
Artificial Intelligence ◽  
Machine Learning ◽  
Data Collection ◽  
Urban Environment ◽  
Evaluation System ◽  
Living Environment ◽  
Mobile Internet ◽  
Practical Significance ◽  
Human Beings ◽  
Urban Living

With the development of urbanization and the continuous development, construction and renewal of the city, the living environment of human beings has also undergone tremendous changes, such as residential community environment and service facilities, urban roads and street spaces, and urban public service formats. And the layout of the facilities, etc., and these are the real needs of people in urban life, but the characteristics of these needs or their problems will inevitably have a certain impact on the user's psychological feelings, thus affecting people's use needs. Then, studying the ways in which urban residents perceive changes in the living environment and how they perceive changes in psychology and emotions will have practical significance and can effectively assist urban management and builders to optimize the living environment of residents. This is also the long-term. One of the topics of greatest interest to urban researchers since then. In the theory of demand hierarchy proposed by American psychologist Abraham Maslow, safety is the basic requirement second only to physiological needs. So safety, especially psychological security, has become one of the basic needs of people in the urban environment. People's perception of the psychological security of the urban environment is also one of the most important indicators in urban environmental assessment. In the past, due to the influence of technical means, the study of urban environmental psychological security often relied on the limited investigation of a small number of respondents. Low-density data is difficult to measure the perceptual results of universality. With the leaping development of the mobile Internet, Internet image data has grown geometrically over time. And with the development of artificial intelligence technology in recent years, image recognition and perception analysis based on machine learning has become possible. The maturity of these technical conditions provides a basis for the study of the urban renewal index evaluation system based on psychological security. In addition to the existing urban visual street furniture data obtained through urban big data collection combined with artificial intelligence image analysis, this paper also proposes a large number of urban living environment psychological assessment data collection strategies. These data are derived from crowdsourcing, and the collection method is limited by the development of cost and technology. At present, the psychological security preference of a large number of users on urban street images is collected by forced selection method, and then obtained by statistical data fitting to obtain urban environmental psychology. Security sense training set. In the future, when the conditions are mature, the brainwave feedback data in the virtual reality scene can be used to carry out the machine learning of psychological security, so as to improve the accuracy of the psychological security data.

scholarly journals Verifiable and Interpretable Reinforcement Learning through Program Synthesis

2019 ◽  
Vol 33 ◽  
pp. 9902-9903
Author(s):  
Abhinav Verma
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Neural Networks ◽  
Reinforcement Learning ◽  
Programming Languages ◽  
Formal Methods ◽  
Domain Knowledge ◽  
Policy Search ◽  
Safety Critical ◽  
Symbolic Methods

We study the problem of generating interpretable and verifiable policies for Reinforcement Learning (RL). Unlike the popular Deep Reinforcement Learning (DRL) paradigm, in which the policy is represented by a neural network, the aim of this work is to find policies that can be represented in highlevel programming languages. Such programmatic policies have several benefits, including being more easily interpreted than neural networks, and being amenable to verification by scalable symbolic methods. The generation methods for programmatic policies also provide a mechanism for systematically using domain knowledge for guiding the policy search. The interpretability and verifiability of these policies provides the opportunity to deploy RL based solutions in safety critical environments. This thesis draws on, and extends, work from both the machine learning and formal methods communities.

scholarly journals Safety Analysis of Deep Neural Networks

2021 ◽  
Author(s):  
Dario Guidotti
Keyword(s):  
Machine Learning ◽  
Neural Networks ◽  
Computer Science ◽  
Deep Neural Networks ◽  
Safety Analysis ◽  
Learning Models ◽  
Safety Critical ◽  
Machine Learning Models

Deep Neural Networks (DNNs) are popular machine learning models which have found successful application in many different domains across computer science. Nevertheless, providing formal guarantees on the behaviour of neural networks is hard and therefore their reliability in safety-critical domains is still a concern. Verification and repair emerged as promising solutions to address this issue. In the following, I will present some of my recent efforts in this area.

scholarly journals Challenges of Machine Learning Applied to Safety-Critical Cyber-Physical Systems

2020 ◽  
Vol 2 (4) ◽  
pp. 579-602
Author(s):  
Ana Pereira ◽  
Carsten Thomas
Keyword(s):  
Machine Learning ◽  
Autonomous Driving ◽  
Cyber Physical Systems ◽  
Shop Floor ◽  
Safety Engineering ◽  
Safety Hazards ◽  
Physical Systems ◽  
Safety Critical ◽  
Holistic Approaches ◽  
Transportation Vehicle

Machine Learning (ML) is increasingly applied for the control of safety-critical Cyber-Physical Systems (CPS) in application areas that cannot easily be mastered with traditional control approaches, such as autonomous driving. As a consequence, the safety of machine learning became a focus area for research in recent years. Despite very considerable advances in selected areas related to machine learning safety, shortcomings were identified on holistic approaches that take an end-to-end view on the risks associated to the engineering of ML-based control systems and their certification. Applying a classic technique of safety engineering, our paper provides a comprehensive and methodological analysis of the safety hazards that could be introduced along the ML lifecycle, and could compromise the safe operation of ML-based CPS. Identified hazards are illustrated and explained using a real-world application scenario—an autonomous shop-floor transportation vehicle. The comprehensive analysis presented in this paper is intended as a basis for future holistic approaches for safety engineering of ML-based CPS in safety-critical applications, and aims to support the focus on research onto safety hazards that are not yet adequately addressed.

Sign in / Sign up

Export Citation Format

Share Document