Novel Techniques for Quantifying Secondhand Smoke Diffusion into Children's Bedroom

<p>The impact of secondhand smoking to health of general population, and specifically children, is a well-known phenomenon that researchers have studied for years. In this dissertation, I extend work done within a secondhand smoke intervention to understand and quantity the impact of intervention on flow of smoke air particle concentration from the main room where smoking generates air particle contamination to a room where a child living in the home sleeps. The paper also explores potential modelling techniques to proactively identify and the impact of the smoke air particles with the intent to discourage adults from smoking in the home and thus potentially minimizing the impact to children’s health. The data was analyzed using hierarchical linear models to quantify the impact of intervention. The analysis finds that smoke air particles attenuated, on average, by 31.6% from the main room to the child’s room. Using hierarchical linear models, I also quantified the effect of intervention where the relationship between the main room and child’s room concentrations decreased once the intervention became active (-0.146 to -0.034 based on random slope versus random intercept). I also developed an LSTM model that can proactively identify whether a smoking event would be an impact children’s health. The results of the model are very encouraging, with an accuracy of approximately 80% when using less than 4 minutes of main room data. The two key outcomes from this study are 1) I can quantify the impact of intervention on the flow of air particle concentration between the main room and child’s room and 2) I am able to develop a modelling approach that can proactively identify the potential impact of SHS to health of the child. The study open doors for several possibilities including use of the findings by practitioners in counselling sessions to provide metrics to smoking adults and advice on the potential impact of smoking to the health of the child. The modelling approach also lays a foundation for future research to implement proactive, real time monitoring and notification in smart homes.</p>

Name thesis:degree_name

Doctor of Philosophy (PhD)

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Computational and Data Sciences

Year

2022

Author dc:creator

• Ramchandani, Sunil

Contributors dc:contributor

• Dr. Vincent Berardi Ph.D
• Dr. Melbourne Hovell Ph.D
• Dr. Cyril Rakowski Ph.D

• Computational Engineering

Repository record dc:identifier

https://digitalcommons.chapman.edu/cads_dissertations/29

OAI identifier oai:identifier

oai:digitalcommons.chapman.edu:cads_dissertations-1029