System and method for determining a sufficient cause from multiple outcomes

A cause-observation technique applied in the field of systems to determine sufficient causes under circumstances

Active Publication Date: 2021-05-07
BAIDU USA LLC
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

If a patient describes a fever, the question becomes: "How can a doctor determine if a patient has pneumonia or the flu or something else?" is a sufficient condition to trigger an alarm; therefore, the question might be asked: "How do I determine if this is a false alarm?

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • System and method for determining a sufficient cause from multiple outcomes
  • System and method for determining a sufficient cause from multiple outcomes
  • System and method for determining a sufficient cause from multiple outcomes

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0031] Example 1 - Variational Inference and then Exact Inference: In an embodiment, a variational inference process can be used to compute the posterior, which is used in the exact inference process. As an example, consider the following:

[0032]

[0033] where a variational inference method is used to find the first positive result set that has been identified via the above-assigned index In case of inferred disease d i (d i represents the probability of the i-th disease).

[0034] In an embodiment, the posterior of the variational inference process may be used as input for exact inference. As an example, consider the following:

[0035]

[0036] where the exact inference method is used for a given second positive result set that has been identified via the assigned index as described above Posterior and Negative Observation Sets F from Variational Inference - Inferred disease in case of d i The probability.

[0037] It should be noted that in an embodiment,...

Embodiment 2

[0038] Example 2 - Exact Inference and Then Variational Inference: In an embodiment, the exact inference process can be used to compute the posterior, which is used in the variational inference process. As an example, consider the following:

[0039]

[0040] where the exact inference method is used to find the second set of positive results and the set of negative results F – Inferred disease in case of d i The probability.

[0041] In an embodiment, the exact inferred posterior may be used as input for variational inference. As an example, consider the following:

[0042]

[0043] where a variational inference method is used to generate the first set of positive results that have been identified via the indices assigned as described above and infer the disease d in the case of the posterior from the exact inference i The probability.

[0044] In an embodiment, in step 225, the results of the inference process are used to output the most likely cause. In the e...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The present invention relates to systems and methods for determining sufficient cause from multiple outcomes for inferring the most likely cause from observable outcomes and known Noisy-OR causality. In an embodiment, the results are sorted by index according to an order including but not limited to natural frequency order of the results, order of expert marks, order obtained by machine learning, and the like. In an embodiment, according to their assigned indices, observations with lower indices are assigned for exact inference, while observations with higher indices are assigned for variational inference. In an embodiment, the results of exact inference and variational inference are combined to predict the most likely cause. The unique combination of exact and variational inference based on outcome indexing makes the process of inferring possible causes much faster.

Description

technical field [0001] The present invention generally relates to providing systems and methods for assisting in determining a sufficient cause given multiple outcomes. Background technique [0002] Causality is the principle that there is a relationship between a cause and an effect or outcome. In some cases, the result may be the result of one of many causes. Various models and theories exist for attempting to formalize causality. One such ensemble of models is called independence of causal influence (ICI), which addresses the problem of exponential growth of parameters when dealing with conditional probabilities by assuming independence of causal influence (ICI). Accepting this assumption allows defining parametric models of conditional probability distributions using only parameters that are linear in multiple causes. [0003] ICI models such as Noisy-OR and Noisy-AND gates have been widely used. The Noisy-OR model is a causal independent formalism that models relati...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Patents(China)
IPC IPC(8): G06N5/04G06N20/00
CPCG06N7/01
Inventor 谢于晟杜楠翟静朱伟铖周达文范伟
Owner BAIDU USA LLC
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap