📅 Date: Thursday, June 12, 9 AM – 10 AM (Hanoi, GMT +7)
🛎 Format: Offline
📍 Venue: VinUniversity, VISHC Open Space

📣 Topic:
1️⃣ DIRECT MEASUREMENT OF THE PLQY ENHANCEMENT OF PEROVSKITE CSPBI3 QUANTUM DOTS FILMS COUPLED WITH A SUB-WAVELENGTH SI3N4 GRATING

📌 Speaker: Ms Le Huyen Trang, Research Assistant, VISHC
💡 Abstract: Perovskite quantum dots (QDs) are attractive for optoelectronic applications, but their efficiency is often limited by non-radiative losses and poor light extraction. In this seminar, we present a strategy to enhance the photoluminescence (PL) and photoluminescence quantum yield (PLQY) of perovskite QDs using photonic crystal (PC) substrates. By coupling QDs to the optical modes of a nanostructured PC, we achieve significant improvements in PL intensity and PLQY. Our results highlight the role of photonic design in tuning light-matter interactions and suggest a scalable approach for improving QD-based devices, such as LEDs and sensors.

2️⃣ COUNTERFACTUAL OUTCOME ESTIMATION OVER TIME FOR LONGITUDINAL DATA UNDER TIME-VARYING CONFOUNDING

📌 Speaker: Mr Nguyen Dai Nghia, PhD Student, UIUC
💡 Abstract: Estimating counterfactual outcomes from observational data, such as electronic health records, holds significant value in many areas, including personalized medicine. A major challenge that prevents straightforward estimation of outcomes from the data is the existence of time-dependent confounding, where covariates affected by past treatments influence future treatment assignments. Recent methods leverage representation learning techniques to learn a balanced representation for bias mitigation. However, these methods may suffer from the risk of over-balancing, which deteriorates covariates distribution and impairs the estimation quality. We propose a novel method for improving the representation quality and enhancing counterfactual estimation performance.