Predicting impact of pollen on allergy severity involves training Bayesian network model on available data for good performance, handling missing data by probabilistic inference, and testing model on available data set by applying log-likelihood and Bayesian Information Criterion
	SHANVI A; SHAIK A M; NAROOKA P; SIMRAN; GUPTA A
	2025-04-04
专利权人	UNIV MANIPAL (UYMA-Non-standard)
申请日期	2025-04-04
专利号	IN202511033226-A
成果简介	NOVELTY - Predicting the impact of pollen on allergy severity involves (a) collecting the data from the sources, such as pollen data is collected from meteorological and allergy monitoring websites, weather data, such as temperature, humidity, wind speed, precipitation from meteorological services, environmental data, such as air pollution levels, allergen dispersion rates, genetic and physiological factors which is susceptibility scores available online, and infection severity data comprising online surveys; (b) converting the data into numeric representations for further processing and normalizing the continuous data; (c) combining the data by designing the Bayesian network model, where all the factors are represented as nodes and there are relationships(edges) drawn out between them; and (d) training Bayesian network model on the 80% of the available data for a good performance. USE - Method for predicting the impact of pollen on allergy severity. ADVANTAGE - The method helps in building more personal predictions as well as management strategies for those who suffer because of such allergies. The method employs Bayesian Networks in identifying the relationship between pollen and allergy severity. The method gives a more personalized touch to the method, which helps predict allergies and makes management strategies on a more personal level. The technology provides great cost advantages by catching high-risk individuals early on before emergency conditions arise, while remaining accessible in the manners of scalability and low computational needs. DETAILED DESCRIPTION - Predicting the impact of pollen on allergy severity involves (a) collecting the data from the sources, such as pollen data is collected from meteorological and allergy monitoring websites, weather data, such as temperature, humidity, wind speed, precipitation from meteorological services, environmental data, such as air pollution levels, allergen dispersion rates, genetic and physiological factors which is susceptibility scores available online, and infection severity data comprising online surveys; (b) converting the data into numeric representations for further processing and normalizing the continuous data; (c) combining the data by designing the Bayesian network model, where all the factors are represented as nodes and there are relationships(edges) drawn out between them; (d) training Bayesian network model on the 80% of the available data for a good performance; (e) handling missing data by probabilistic inference; and (f) testing the model on 20% of the available data set by applying log-likelihood and Bayesian Information Criterion (BIC).
IPC 分类号	A61K-039/36 ; A61P-037/08 ; E21B-049/00 ; G06N-020/00 ; G06N-007/01
国家	印度
专业领域	信息技术
语种	英语
成果类型	专利
文献类型	科技成果
条目标识符	http://119.78.100.226:8889/handle/3KE4DYBR/13200
专题	中国科学院新疆生态与地理研究所
作者单位	UNIV MANIPAL (UYMA-Non-standard)
推荐引用方式 GB/T 7714	SHANVI A,SHAIK A M,NAROOKA P,et al. Predicting impact of pollen on allergy severity involves training Bayesian network model on available data for good performance, handling missing data by probabilistic inference, and testing model on available data set by applying log-likelihood and Bayesian Information Criterion. IN202511033226-A[P]. 2025.

条目包含的文件
条目无相关文件。

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。