Search

BACKGROUND Previous research has shown the feasibility of utilizing social media data from a singular platform (e.g., Facebook or Twitter) in distinguishing individuals with a diagnosis of mental illness or experiencing an adverse outcome from healthy volunteers. However, the performance of these models on data from other social media platforms unseen in the training data (e.g., Instagram, TikTok) have not been investigated. OBJECTIVE This study aims to explore if online identities fragmented across social media platforms, models would have better testing performance on data from already seen social media platforms, in comparison to unseen social media platforms. It also aims to explain such discrepancies in performances if they are found. METHODS Windowed timeline data from three platforms with clinically-verified labels of hospitalization among patients with a diagnosis of schizophrenia was gathered: Facebook (N = 254), Twitter (N = 54), and Instagram (N = 124). Then, we utilized a 3 x 3 combinatorial binary classification design to test model’s performance on testing data from all available platforms. We further compared results from models within intra-platform experiments (i.e., training and testing data belongs to the same platform) to models within inter-platform experiments (i.e., training and testing data belongs to the different platforms). Finally, we utilized SHapley Additive exPlanations (SHAP) to extract top predictive features to explain the underlying constructs that predict hospitalization on each platform. RESULTS We found that models within intra-platform experiments on average achieved an F1-score of 0.72 in detection a psychiatric hospitalization due to schizophrenia, which is 68% higher compared to the average of models within inter-platform experiments at an F1-score of 0.428. We also found that by combining training data of all three platforms, a slight improvement of 0.5% was observed on the testing sets on average, compared to original intra-platform models. An analysis of top features for the intra-platform models shows low predictive feature overlap between the platforms, with ‘anger’ being an unique top feature for Facebook while ‘sad’ being an unique top feature for Instagram. CONCLUSIONS We demonstrated models built on one platform’s data to predict critical mental health treatment outcomes, such as a hospitalization, may not generalize to another, because each platform offers different construct validity. However, combining data from multiple platforms together may offer a more comprehensive view of a patient’s state and situation, and therefore fare better in relapse prediction. With the changing ecosystem of social media use among different demographic groups and as online identities continue to get fragmented across platforms, further research on holistic approaches to harnessing these diverse data sources is required.

Measuring the similarity between materials is essential for estimating their properties and revealing the associated physical mechanisms. However, current methods for measuring the similarity between materials rely on theoretically derived descriptors and parameters fitted from experimental or computational data, which are often insufficient and biased. Furthermore, outliers and data generated by multiple mechanisms are usually included in the dataset, making the data-driven approach challenging and mathematically complicated. To overcome such issues, we apply the Dempster–Shafer theory to develop an evidential regression-based similarity measurement (eRSM) method, which can rationally transform data into evidence. It then combines such evidence to conclude the similarities between materials, considering their physical properties. To evaluate the eRSM, we used two material datasets, including 3[Formula: see text] transition metal–4[Formula: see text] rare-earth binary and quaternary high-entropy alloys with target properties, Curie temperature, and magnetization. Based on the information obtained on the similarities between the materials, a clustering technique is applied to learn the cluster structures of the materials that facilitate the interpretation of the mechanism. The unsupervised learning experiments demonstrate that the obtained similarities are applicable to detect anomalies and appropriately identify groups of materials whose properties correlate differently with their compositions. Furthermore, significant improvements in the accuracies of the predictions for the Curie temperature and magnetization of the quaternary alloys are obtained by introducing the similarities, with the reduction in mean absolute errors of 36% and 18%, respectively. The results show that the eRSM can adequately measure the similarities and dissimilarities between materials in these datasets with respect to mechanisms of the target properties.

We present a data-driven approach to explore high-entropy alloys (HEAs). To overcome the challenges with numerous element-combination candidates, selecting appropriate descriptors, and the limitations and biased of existing data, we apply the evidence theory to develop a descriptor-free evidence-based recommender system (ERS) for recommending HEAs. The proposed system measures the similarities between element combinations and utilizes it to recommend potential HEAs. To evaluate the ERS, we compare its HEA-recommendation capability with those of matrix-factorization- and supervised-learning-based recommender systems on four widely known data sets, including binary and ternary alloys. The results of experiments using k-fold cross-validation on the data sets show that the ERS outperforms all competitors. Furthermore, the ERS shows excellent extrapolation capabilities in experiments of recommending quaternary and quinary HEAs. We experimentally validate the most strongly recommended Fe-Co-based magnetic HEA, viz. FeCoMnNi, and confirm that it shows a body-centered cubic structure and is stable at high temperatures.

The thesis demonstrates for the first time coexistence of write once read many memory (WORM) with resistive random access memory (RRAM). This is the first time such coexistence phenomena are reported in organic materials system. Furthermore, better understanding of working mechanism in resistive switching phenomena in organic materials particularly conjugated polymer has been achieved. Two materials systems are chosen, namely, poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT PSS) and Polyvinyl alcohol (PVA) where PEDOT PSS is a conducting polymer and PVA is an insulator polymer. For a device with 5 μm thick PEDOT PSS sandwiched between two Au electrodes, the coexistence of WORM memory and RRAM memory can be achieved (Chapter 4). The WORM result was attributed to dedoping of the conducting PEDOT + PSS- segments to insulating PEDOT 0 segments. The RRAM was explained due to Proton H+ conductive properties of PSS acid domains inside the film after dedoping. This points to a general feature: resistive switching behavior of a whole device is the result of additive/summative resistive switching characteristic of its components. Based on this, we reveal the coexistence of RRAM memory and WORM memory demonstrated in PEDOT PSS blend with PVA (Chapter 5). The WORM memory is again attributed to PEDOT PSS while RRAM memory is attributed to the charge trapping characteristic of PVA. The device performance is improved significantly as compared to the earlier PEFOT PSS system, with more than 1000 cycling stability and much more than 10^4 s retention for RRAM mode. Additional I-V characteristic of this device is further studied in chapter 6 which reveals a new type of WORM memory. This type of WORM memory is driven by metallic migration of Au electrodes. Doctor of Philosophy (MSE)

In this paper, we introduce an ECHONET Lite based IoT platform which allows households to set up smart home environment with ease as well as allows service providers to deploy their IoT applications easily and securely. Our proposed platform is fully implemented with three main features: a simple process for integrating new IoT devices into smart home environment, a seamless data transfer mechanism between IoT devices and service providers, and a RESTful API for verifying and supporting service providers to deploy their IoT applications into our platform. The experimental results verified the practicability of our proposed IoT platform.

Social security is one of the most important issues in socio-economic manage-ment, especially in the current stage of integration and development. In recent years, Vietnam has been highly appreciated by the international community for many achievements in economic development associated with poverty reduction and social security. In the structure of capital for social welfare goals in Vietnam, capital gained through ODA plays an important role. However, the attraction of ODA capital to Vietnam in recent years still faces many difficulties. By using qualitative and quantitative research methods, this paper examines the impact of the factors affecting the attraction of ODA capital into Vietnam and offers possi-ble solutions to increase this source by the end of 2025. Depth interviews through open-ended questions were used to collect data from 06 Vietnamese or-ganizations receiving ODA and 05 institutions representing ODA sponsors such as WB; UNECF; USAIS; JICA… Basing on their answers, the authors contin-ued to ask questions to better exploit the factors affecting ODA attraction. The interviews were recorded and compared with each other to find the similarities or differences of each organizations and institutions. Consequently, the authors build measurement scales, questionnaires, and propose research models and hy-potheses. The research’s results show that the efficiency of using ODA is the most appreciated factor. It has a great impact on the ability to attract ODA to im-plement the goal of ensuring social security in Vietnam: when this factor is in-creased to 1 unit, the ability to attract ODA will increase to 0.222 units. Regres-sion analysis results of policies, strategies, institutions and legislations also sup-port the attraction of ODA. The target conformity between the donor and the recipient of ODA has a positive relationship with the attraction of ODA.

The study assesses the impact of different types of land use rights certificates (LURC) on individual and household welfare, expanding on the previous research of Menon, Rodgers, and Kennedy (2016), which assessed the effects of LURCs on agricultural land on household welfare. This study considers more recent data from the Vietnam Household Living Standards Survey (VHLSS) of 2014 in addition to VHLSS data from 2002–08 and includes an analysis of LURCs for agricultural as well as residential land. The study findings are presented below in three sections. The first is an analysis of land use and LURC distribution trends based on the VHLSS data. The second draws on a group of impact evaluations that compares the effects of having different types of LURCs. We regress individuals' employment and health care outcomes on whether their name was included on a LURC. The treatment variables are whether one has been issued an agricultural or residential LURC, with controls for age, education, ethnicity, urban residence, household consumption, land area, and district. We also regress household-level outcomes—expenditures, credit levels, and incomes—on whether the household's LURC is singly or jointly titled. Notably, the distribution of LURCs is not randomized, making it difficult to estimate the causal effects of LURC status and thereby limiting the conclusions that can be drawn. The third section, using 2014 demographic data and LURC distributions, offers a cost-benefit analysis of efforts to convert and reissue all remaining singly titled LURCs to jointly titled ones. We estimate the benefit as the impact difference between single- and jointly titled LURCs as calculated in the impact evaluation section and estimate costs as those of reissuing a LURC.

Molybdenum trioxide is an interesting inorganic system in which the empty 4d states have potential to hold extra electrons and therefore can change states from insulating opaque (MoO3) to colored semimetallic (HxMoO3). Here, we characterize the local electrogeneration and charge transfer of the synthetic layered two-dimensional 2D MoO3-II (a polymorph of MoO3 and analogous to α-MoO3) in response to two different redox couples, i.e., [Ru(NH3)6]3+ and [Fe(CN)6]3- by scanning electrochemical microscopy (SECM). We identify the reduction of [Ru(NH3)6]3+ to [Ru(NH3)6]2+ at the microelectrode that leads to the reduction of MoO3-II to conducting blue-colored molybdenum bronze HxMoO3. It is recognized that the dominant conduction of the charges occurred preferentially at the edges active sites of the sheets, as edges of the sheets are found to be more conducting. This yields positive feedback current when approaching the microelectrode toward 2D MoO3-II-coated electrode. In contrast, the [Fe(CN)6]4-, which is reduced from [Fe(CN)6]3-, is found unfavorable to reduce MoO3-II due to its higher redox potential, thus showing a negative feedback current. The charge transfer on MoO3-II is further studied as a function of applied potential. The results shed light on the charge transfer behavior on the surface of MoO3-II coatings and opens the possibility of locally tuning of their oxidation states. National Research Foundation (NRF) We acknowledge financial support from the National Research Foundation Investigatorship Award, NRF-NRFI-2016-05. Part of this work was also supported by the Nanomaterials for Energy and Water Management Programme under the Campus for Research Excellence and Technological Enterprise (CREATE) that is supported by the National Research Foundation, Prime Minister’s Office, Singapore.

We investigate determinants of individual migration decisions in Vietnam, a country with increasingly high levels of geographical labour mobility. Using data from the Vietnam Household Living Standards Survey 2012 (VHLSS2012), we find that the probability of migration is strongly associated with individual, household and community-level characteristics. The probability of migration is higher for young people and those with post-secondary education. Migrants are more likely to be from households with better-educated household heads, female-headed households, and households with higher youth dependency ratios. Members of ethnic minority groups are much less likely to migrate, other things being equal. Using multinomial logit methods, we distinguish migration by broad destination, and find that those moving to Ho Chi Minh City or Hanoi have broadly similar characteristics and drivers of migration as those moving to other destinations. We also use the VHLSS2012 together with the VHLSS2010, which allows us to focus on a narrow cohort of recent migrants—those present in the household in 2010, but who had moved away by 2012. This yields much tighter results. For education below upper secondary school, the evidence on positive selection by education is much stronger. However, the ethnic minority “penalty” on spatial labor mobility remains strong and significant, even after controlling for specific characteristics of households and communes. This lack of mobility is a leading candidate to explain the distinctive persistence of poverty among Vietnam’s ethnic minority populations, even as national poverty has sharply diminished.

Transparent nonvolatile memory has great potential in integrated transparent electronics. Here, we present highly transparent resistive switching memory using stoichiometric WO3 film produced by cathodic electrodeposition with indium tin oxide electrodes. The memory device demonstrates good optical transmittance, excellent operative uniformity, low operating voltages (+0.25 V/–0.42 V), and long retention time (>104 s). Conductive atomic force microscopy, ex situ transmission electron microscopy, and X-ray photoelectron spectroscopy experiments directly confirm that the resistive switching effects occur due to the electric field-induced formation and annihilation of the tungsten-rich conductive channel between two electrodes. Information on the physical and chemical nature of conductive filaments offers insightful design strategies for resistive switching memories with excellent performances. Moreover, we demonstrate the promising applicability of the cathodic electrodeposition method for future resistive ...

Self-powered flexible graphene thermistor device is successfully fabricated by modifying the graphene layers with ferroelectric polarization. Chemical vapor deposited graphene monolayer was transferred to the surface of a ferroelectric polymer poly[(vinylidenefluoride- co -trifluoroethylene)] P(VDF-TrFE) thin film. By utilizing alternating polarization charge, the graphene surface contains polarization induced hole-rich, intrinsic and electron-rich regions (p-i-n regions). The potential difference created in the graphene layer leads to the drift of the extra carriers created in the graphene monolayer with increasing temperature and gives rise to an exemplary self-powered or autonomous thermistor.

Hexagonal boron nitride (hBN), which is a 2D layered dielectric material, sometimes referred as “white graphene” due to its structural similarity with graphene, has attracted much attention due to its fascinating physical properties. Here, for the first time the use of chemical vapor deposition -grown hBN films to fabricate ultrathin (≈3 nm) flexible hBN-based resistive switching memory device is reported, and the switching mechanism through conductive atomic force microscopy and ex situ transmission electron microscopy is studied. The hBN-based resistive memory exhibits reproducible switching endurance, long retention time, and the capability to operate under extreme bending conditions. Contrary to the conventional electrochemical metallization theory, the conductive filament is found to commence its growth from the anode to cathode. This work provides an important step for broadening and deepening the understanding on the switching mechanism in filament-based resistive memories and propels the 2D material application in the resistive memory in future computing systems.

Miniaturization of electronic devices to gain speed and reduce cost is no longer the only criterion for emerging technological needs. The devices with advanced functionalities such as flexibility, stretchability, transparency, and environmental robustness have significant advantages and added value in different areas for future electronics' applications. Data storage devices are one of the critical and fundamental components in fully integrated electronics. The functional resistive switching random access memory (RRAM) device, which is one of the most promising candidates for the next generation of non-volatile memory for data storage, has attracted immense attention because of its speed, ease of fabrication and scalability. In this paper, the recent progress, materials' selection, device architecture, and the emerging multi-functional RRAM devices are reviewed.

We analyze Curie temperatures of rare-earth transition metal binary alloys with machine learning method. In order to select important descriptors and descriptor groups, we introduce newly developed subgroup relevance analysis and adopt the hierarchical clustering in the representation. We execute the exhaustive search and successfully illustrate the importance of descriptors and descriptor groups. We execute the exhaustive search and illustrate that our approach indeed leads to the successful selection of important descriptors and descriptor groups. It helps us to choose the combination of the descriptors and to understand the meaning of the selected combination of descriptors., 16 pages, 8 figures

In this paper, a nanocrystalline porous silicon-based propulsion system is designed, and its explosion impulse is tuned by varying the propulsion design parameters and etching duration. The porous silicon is ignited by an electrical current passed through 100-nm-thick aluminum film deposited on the unpolished side of the wafer and the ignited porous silicon caused strong explosion, which destroys the chip into tiny fragments. The explosion impulse in the system can reach about 0.14 N·s at optimal conditions, which is two orders stronger than the impulse produced by conventional propellants (ZakarE., “Technology Challenges in Solid Energetic Materials for Micro Propulsion Applications” U.S. Army Research Lab. Rept. ARL-TR-5035, Nov. 2009). It is also shown that, by varying the etching time, which is an important factor that determines the porous layer thickness and the volume of nanocrystallite, the strength of the impulse can be tuned. Furthermore, a linear increasing trend of the explosion impulse with ...