Dataset Information

Hardware implementation of Bayesian network building blocks with stochastic spintronic devices

ABSTRACT: Bayesian networks are powerful statistical models to understand causal relationships in real-world probabilistic problems such as diagnosis, forecasting, computer vision, etc. For systems that involve complex causal dependencies among many variables, the complexity of the associated Bayesian networks become computationally intractable. As a result, direct hardware implementation of these networks is one promising approach to reducing power consumption and execution time. However, the few hardware implementations of Bayesian networks presented in literature rely on deterministic CMOS devices that are not efficient in representing the stochastic variables in a Bayesian network that encode the probability of occurrence of the associated event. This work presents an experimental demonstration of a Bayesian network building block implemented with inherently stochastic spintronic devices based on the natural physics of nanomagnets. These devices are based on nanomagnets with perpendicular magnetic anisotropy, initialized to their hard axes by the spin orbit torque from a heavy metal under-layer utilizing the giant spin Hall effect, enabling stochastic behavior. We construct an electrically interconnected network of two stochastic devices and manipulate the correlations between their states by changing connection weights and biases. By mapping given conditional probability tables to the circuit hardware, we demonstrate that any two node Bayesian networks can be implemented by our stochastic network. We then present the stochastic simulation of an example case of a four node Bayesian network using our proposed device, with parameters taken from the experiment. We view this work as a first step towards the large scale hardware implementation of Bayesian networks.

SUBMITTER: Debashis P

PROVIDER: S-EPMC7524796 | biostudies-literature | 2020 Jan

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Background:The Six Building Blocks for improving opioid management (6BBs) is a program for improving the management of patients in primary care practices who are on long-term opioid therapy for chronic pain. The 6BBs include building leadership and consensus; aligning policies, patient agreements, and workflows; tracking and monitoring patient care; conducting planned, patient-centered visits; tailoring care for complex patients; and measuring success. The Agency for Healthcare Research and Quality funded the development of a 6BBs implementation guide: a step-by-step approach for independently implementing the 6BBs in a practice. This mixed-method study seeks to assess practices' use of the implementation guide to implement the 6BBs and the effectiveness of 6BBs implementation on opioid management processes of care among practices using the implementation guide. Methods:Data collection is guided by the Consolidated Framework for Implementation Research, Proctor's taxonomy of implementation outcomes, and the Centers for Disease Control and Prevention's Guideline for Prescribing Opioids for Chronic Pain. A diverse group of health care organizations with primary care clinics across the USA will participate in the study over 15?months. Qualitative data collection will include semi-structured interviews with stakeholders at each organization at two time points, notes from routine check-in calls, and document review. These data will be used to understand practices' motivation for participation, history with opioid management efforts, barriers and facilitators to implementation, and implementation progress. Quantitative data collection will consist of a provider and staff survey, an implementation milestones assessment, and quarterly opioid prescribing quality measures. These data will supplement our understanding of implementation progress and will allow us to assess changes over time in providers' opioid prescribing practices, prescribing self-efficacy, challenges to providing guideline-driven care, and practices' opioid prescribing quality measures. Qualitative data will be coded and analyzed for emergent themes. Quantitative data will be analyzed using descriptive statistics and clustered multivariate regression. Discussion:This study contributes to the knowledge of the implementation and effectiveness of a team-based approach to opioid management in primary care practices. Information gleaned from this study can be used to inform efforts to curtail opioid prescribing and assist primary care practices considering implementing the 6BBs.

Dataset Information

Hardware implementation of Bayesian network building blocks with stochastic spintronic devices

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets