Suchergebnisse

This paper analyzes the Catholic contributions to the development of the Vietnamese alphabet—quốc ngữ, or the "national script"—during its first 250 years of existence, before it was popularized outside of Catholic circles. In particular, the article traces the evolution of quốc ngữ spelling in the writings and publications of European Catholic missionaries and Vietnamese catechists and priests from the seventeenth to the nineteenth centuries. These texts include the dictionaries and descriptions of the Vietnamese language by Alexandre de Rhodes (1651) and Jean-Louis Tabert (1838), as well as lesser-known quốc ngữ writings before the twentieth century. Since the Vietnamese alphabet began as a transcription of spoken Vietnamese, its changing orthography also reflects changes in Vietnamese pronunciation.

AbstractThis study analyzes the influence of children on household migration decisions using data on current internal movement in Vietnam a country that has experienced significant rural–urban migration in the recent years. Families with children usually have three migration choices: move together, stay together or send only one parent to work afar. Using an instrumental variable approach, we show that having an additional child reduces the probability of household migration by 0.0115, while it increases the likelihood of fathers' migration by 0.0121. These effects suggest that households with more children may be less mobile but may have a greater economic need for migration.

International audience ; L'équilibrage de charge consisteà répartir le trafic entre une paire de noeuds sur plusieurs chemins afin d'améliorer les performances du réseau. Idéalement, la politique d'équilibrage devrait s'adapter en fonction du trafic et anticiper son impact sur le réseau mais l'intégration de modèles pour la Qualité de Service est un défi. Nous proposons une solution basée sur l'apprentissage par renforcement profond, qui est capable d'apprendre la relation entre le trafic et la QoS, tout en offrant la sécurité nécessaire pour maximiser le débit etéviter de violer les contraintes de capacité des liaisons. Elle intègre un algorithme d'optimisation de politiques sous contraintes. Dans un scénario SD-WAN où les délais suivent le modèle de mise en file d'attente M/M/1, nous démontrons,à l'aide d'un programme non linéaire en nombres entiers, que notre solution peut atteindre un délai de bout en bout proche de l'optimum. Nous montronségalement que notre solution apprend automatiquement les paramètres de récompense pour répondre aux contraintes de capacité. Mots-clefs : Equilibrage de charge, apprentissage par renforcement, SD-WAN.

Abstract. We investigate the most severe rainfall event ever experienced in Ho Chi Minh City (HCMC), Vietnam. It occurred on 25 November 2018 when Typhoon (TY) Usagi directly hit HCMC. During this event, there was more than 300 mm of rainfall over 24 h which led to flooding and considerable material damage. We propose an in-depth study of TY-induced, compound flood drivers at a short timescale by focusing on the days before and after the event. We use a set of data analysis and signal processing tools to characterize and quantify both coastal and inland effects on the hydrosystem. We found that TY Usagi made landfall without forming a significant storm surge. The extreme rainfall does not translate into immediate river discharge but presents a 16 h time lag between peak precipitation and peak residual discharge. Nevertheless, increased river water levels can be seen at both urban and upstream stations with a similar time lag. At the upstream river station, residual discharge represents 1.5 % of available rainwater, and evidence of upstream widespread flooding was found. At the urban river station, we assess the potential surface runoff during the event to be 8.9 % of the upstream residual discharge. However, a time lag in peak river water level and peak rainfall was found and attributed to the combination of high tide and impervious streets which prevented the evacuation of rainwater and resulted in street flooding of up to 0.8 m. Overall, it was found that despite not having a significant storm surge, the coastal tidal forcing is the predominant compound flood driver even during severe, heavy rainfall with tidal fluctuations in river water level and respective discharge much larger than the residuals.