China
Africa
Explore chapters and articles related to this topic
Introduction to Stochastic Decision-Making Models for Business Analytics
Published in William P. Fox, Mathematical Modeling for Business Analytics, 2017
What is stochastic modeling? Stochastic modeling concerns the use of probability to model real-world situations in which uncertainty is present. As uncertainty is pervasive, this means that the tools that we present can potentially prove useful in almost all facets of one’s professional life and in one’s personal life. Such topics are as follows: GamblingPersonal or business financesDisease treatment optionsEconomic forecastingProduct demandCall center provisioningProduct reliability and warranty analysis, and so on
Introduction
Published in Raymond J. Madachy, Daniel X. Houston, What Every Engineer Should Know About Modeling and Simulation, 2017
Raymond J. Madachy, Daniel X. Houston
A simulation model can be deterministic, stochastic, or hybrid (mixed). In the deterministic case, all input parameters are specified as single values. Stochastic modeling recognizes the inherent uncertainty in many parameters and relationships using random numbers drawn from a specified probability distributions. Mixed modeling employs both deterministic and stochastic parameters.
CHIMERA Fusion Technology Facility: Testing and Virtual Qualification
Published in Fusion Science and Technology, 2023
Thomas R. Barrett, M. Bamford, N. Bowden, B. Chuilon, T. Deighan, P. Efthymiou, M. Gorley, T. Grant, D. Horsley, M. Kovari, M. Tindall
The virtual design and qualification approach can also account for the uncertainty that must inevitably be managed (and communicated to stakeholders) in the absence of physical testing under real reactor conditions. Stochastic modeling involves treating every input variable to a simulation as a parameter distribution and reporting every output also with a probability distribution. Measured outputs (from CHIMERA or another operational plant) and simulated outputs are utilized together to provide updated model parameters and determine the most probable machine state. Examples of known uncertainties for which we can input a quantified (engineered) distribution are component fatigue life (material s-n data scatter), manufacturing tolerances, surface roughness, joint friction coefficient, or joint contact resistance. Advances in computing hardware, and the systems simulation approach, make it possible to run large numbers of simulations concurrently to explore the parameter distributions. Ultimately, the lack of a fully prototypic test facility means that end-of-life fusion component data are not obtainable, but this approach means that neither should they be required. Virtual assessments with uncertainty quantification enable more effective design, and the risk to in-vessel components is more likely to be made palatable to operators and investors.
Development of Large-Scale Farming Based on Explainable Machine Learning for a Sustainable Rural Economy: The Case of Cyber Risk Analysis to Prevent Costly Data Breaches
Published in Applied Artificial Intelligence, 2023
The scientific innovation of the proposed framework lies in the integration of explainable machine learning (exML) techniques into large-scale farming to enhance cyber risk analysis and mitigation. While the agricultural industry has witnessed the transformation brought about by technology, including machine learning and advanced tools, it has also become vulnerable to cyber risks and data breaches. Therefore, the framework aims to address this specific challenge by introducing novel methodologies and approaches. Incorporating exML Techniques: The integration of explainable machine learning techniques into the framework allows for transparent and interpretable risk analysis. By utilizing exML models, the framework can provide insights into the decision-making process of the system, enabling organizations to understand how risks are identified, assessed, and managed. This enhances the overall transparency and trustworthiness of the risk analysis process.Analytical Stochastic Modeling of Risk: The framework introduces analytical stochastic modeling to quantify and assess cyber risks in large-scale farming. This modeling approach takes into account various factors and uncertainties associated with cyber threats, enabling a more comprehensive and realistic risk assessment. By incorporating stochastic modeling, the framework provides a more robust and accurate representation of risk probabilities and potential impacts.Multi-Criteria Objective Function: The framework utilizes a multi-criteria objective function to balance investment value and cost in risk management decisions. This approach considers multiple criteria, such as the value of assets at risk, the potential financial impact of data breaches, and the cost-effectiveness of mitigation measures. By incorporating these criteria into the decision-making process, organizations can make informed choices that optimize resource allocation and risk reduction efforts.Sustainable Rural Economy Focus: The proposed framework highlights the importance of fostering a sustainable rural economy through effective cyber risk management. By safeguarding sensitive agricultural data and protecting against data breaches, the framework contributes to building resilience and maintaining trust in the agricultural industry. This focus on sustainability aligns with broader societal goals of economic stability and longevity in rural areas.