The agriculture sectors in Canada are highly vulnerable to a wide range of inter-related weather risks linked to seasonal climate variability (e.g., El Ni ̃no Southern Oscillation(ENSO)), short-term extreme weather events (e.g., heatwaves), and emergent disease(e.g., grape powdery mildew). All of these weather-related risks can cause severe crop losses to agricultural crop yield and crop quality as Canada grows a wide range of farm products, and the changing weather conditions mainly drive farming practices. This dissertation presents three machine learning-based statistical models to assess the weather risks on the Canadian agriculture regions and to provide reliable risk forecasting to improve the decision-making of Canadian agricultural producers in farming practices. The first study presents a multi-scale, cluster-based Principal Component Analysis(PCA) approach to assess the potential seasonal impacts of ENSO to spring wheat and barley on agricultural census regions across the Canada prairies areas. Model prediction skills for annual wheat and barley yield have examined in multi-scale from spatial cluster approaches. The ’best’ spatial models were used to define spatial patterns of ENSO forcing on wheat and barley yields. The model comparison of our spatial model to non-spatial models shows spatial clustering and ENSO forcing have increase model performance of prediction skills in forecasting future cereal crop production. The second study presents a copula-Bayesian network approach to assess the impact of extreme high-temperature events (heatwave events) on the developments of regional crops across the Canada agricultural regions at the eco-district-scale. Relevantweather variables and heatwave variables during heatwave periods have identified and used as input variables for model learning. Both a copula-Bayesian network and Gaussian-based network modeling approach is evaluated and inter-compared. The copula approach based on ’vine copulas’ generated the most accurate predictions of heatwave occurrence as a driver of crop heat stress. The last study presents a stochastic, hybrid-Bayesian machine-learning approach to explore the complex causal relationships between weather, pathogen, and host for grape powdery mildew in an experimental farm in Quebec, Canada. This study explores a high-performance network model for daily disease risk forecast by using estimated development factors of pathogen and host from recorded daily weather variables. A fungicide strategy for disease control has presented by using the model outputs and forecasted future weather variability. The dissertation findings are beneficial to Canada’s agricultural sector. The inter-related weather risks explored by the three separate studies in multi-scales provide a better understanding of the interactions between changing weather conditions, extreme weather, and crop production. The research showcases new insights, methods, and tools for minimizing risk in agricultural decision-making / Graduate / 2021-08-19

Contexte : Les neuroleptiques sont souvent prescrits chez les sujets âgés et les patients souffrant de schizophrénie qui sont des personnes vulnérables. Notre objectif était d’étudier l’impact des NLP en situation réelle de prescription, dans ces deux populations. Etude 1. Consommation de neuroleptiques et décès en période de canicule chez les sujets âgés. A partir de données de l’Assurance Maladie, nous avons comparé les prescriptions de NLP chez des sujets âgés décédés pendant la canicule d’août 2003 (n=11624) aux prescriptions de témoins non décédés. Nous avons mis en évidence une association entre risque de décès et consommation de neuroleptiques, que ce soit juste avant ou pendant le pic de canicule et indépendamment d’autres médicaments, d’une démence ou d’une pathologie cardiaque. Etude 2. Efficacité réelle des NLP chez les patients souffrant de schizophrénie. A partir de données d’une cohorte observationnelle ayant inclus en France entière des schizophrènes adultes, nous avons montré que chez les patients déjà traités par NLP (n=5500), il y avait une association entre traitement par antipsychotique atypique (vs. NLP classique) et une meilleure satisfaction avec les soins et ce, pour tous les AA pris en compte et indépendamment du niveau de symptomatologie. Par ailleurs, chez les patients naïfs vis-à-vis de tout NLP et pour qui un traitement était introduit pour la première fois (n=467), un tiers des patients ne s’améliorait pas. Les facteurs prédictifs d’une meilleure réponse clinique étaient une moindre sévérité initiale globale des symptômes et des symptômes négatifs de schizophrénie. Au total, il existait cinq types de trajectoires d’évolution clinique. / Context : Antipsychotic drugs (AD) are often prescribed to elderly people and patients with schizophrenia and both populations are fragile. The aim of the present thesis was to investigate the impact of AD in these patients, in a real-life setting. Study 1. Antipsychotic drug use during a heatwave and risk of death, in older people. Using data from the French Social Security, prescriptions of older people who died during the heatwave in August 2003 (n=11624) were compared to these of controls who survived. An association was evidenced between a prescription of AD and death, in the periods just preceding and during the heatwave, independently of other psychotropic drugs, dementia or cardiac disease. Study 2. Effectiveness of AD in schizophrenia patients. Using data from a French nationwide observational study, we found that in patients treated for schizophrenia (n=5500), a higher level of satisfaction was independently associated with being on second-generation AD as compared to first-generation AD. Also, in drug-naive patients (n=467) followed up during 6 months after a first AD initiation, one third of the patients did not experience any improvement of symptoms. The factors predictive of a better clinical response were lower levels of baseline negative symptoms and overall severity of symptoms. A total of 5 trajectories of clinical evolution were identified.