TY - GEN
T1 - Using Bayesian Belief Network to Solve Abnormal Symptoms after Vaccination under Modern Information Technology
AU - Kuan, Sin Kin
N1 - Publisher Copyright:
© 2021 ACM.
PY - 2021/10/23
Y1 - 2021/10/23
N2 - Rare side effects are weakening confidence in the vaccine. The question is how we interpret the data. Within 15 months after the discovery of the new coronavirus, a variety of effective and safe vaccines against the new coronavirus were available. After receiving the new coronavirus vaccine, some people developed facial paralysis, thigh pain, and even cerebral venous thrombosis. Although these side effects are very rare, and there is a lack of clarity whether there is a causal relationship with the vaccine or not, such news may undermine the confidence of the global vaccine. In order to maintain the confidence of the public, adverse events after vaccination are called ordinary events, and deaths occurring within a few days after vaccination are also interpreted as being caused by their latent diseases. From the following research, the issue of causality divides the vaccinated population into healthy groups and long-term patient groups, and use Bayesian belief network to analyze whether there are symptoms or abnormal events after vaccination as well as the probability distribution of rare illness, death, etc., in order to understand the relationship among each other. Therefore, suspending the administration of COVID vaccine is not a zero-risk option. The reality is that nothing is without risk. Measures to mitigate a risk must be balanced with competitive hazards. Risk seems to be an abstract and vague concept. Risk can be reduced, but it can never be eliminated. The advantage of the Bayesian model is that it is easy to bring the data of various variables into the graph and calculate the posterior data from the known data to strengthen the persuasiveness of vaccination. By using Bayesian Network with PGM Module of Pytorch, the death probability of these two groups can be calculated under abnormal symptoms or without them. The simulation result of death after inoculation is lower than that of normal state without Covid-19 pandemic.
AB - Rare side effects are weakening confidence in the vaccine. The question is how we interpret the data. Within 15 months after the discovery of the new coronavirus, a variety of effective and safe vaccines against the new coronavirus were available. After receiving the new coronavirus vaccine, some people developed facial paralysis, thigh pain, and even cerebral venous thrombosis. Although these side effects are very rare, and there is a lack of clarity whether there is a causal relationship with the vaccine or not, such news may undermine the confidence of the global vaccine. In order to maintain the confidence of the public, adverse events after vaccination are called ordinary events, and deaths occurring within a few days after vaccination are also interpreted as being caused by their latent diseases. From the following research, the issue of causality divides the vaccinated population into healthy groups and long-term patient groups, and use Bayesian belief network to analyze whether there are symptoms or abnormal events after vaccination as well as the probability distribution of rare illness, death, etc., in order to understand the relationship among each other. Therefore, suspending the administration of COVID vaccine is not a zero-risk option. The reality is that nothing is without risk. Measures to mitigate a risk must be balanced with competitive hazards. Risk seems to be an abstract and vague concept. Risk can be reduced, but it can never be eliminated. The advantage of the Bayesian model is that it is easy to bring the data of various variables into the graph and calculate the posterior data from the known data to strengthen the persuasiveness of vaccination. By using Bayesian Network with PGM Module of Pytorch, the death probability of these two groups can be calculated under abnormal symptoms or without them. The simulation result of death after inoculation is lower than that of normal state without Covid-19 pandemic.
KW - Bayesian Network
KW - Covid-19
KW - Vaccine
KW - epidemic
UR - http://www.scopus.com/inward/record.url?scp=85126652441&partnerID=8YFLogxK
U2 - 10.1145/3495018.3495337
DO - 10.1145/3495018.3495337
M3 - Conference contribution
AN - SCOPUS:85126652441
T3 - ACM International Conference Proceeding Series
SP - 1064
EP - 1069
BT - Proceedings of 2021 3rd International Conference on Artificial Intelligence and Advanced Manufacture, AIAM 2021
PB - Association for Computing Machinery
T2 - 3rd International Conference on Artificial Intelligence and Advanced Manufacture, AIAM 2021
Y2 - 23 October 2021 through 25 October 2021
ER -
