Approximating Functions with Approximate Privacy for Applications in Signal Estimation and Learning.

Large corporations, government entities and institutions such as hospitals and census bureaus routinely collect our personal and sensitive information for providing services. A key technological challenge is designing algorithms for these services that provide useful results, while simultaneously maintaining the privacy of the individuals whose data are being shared. Differential privacy (DP) is a cryptographically motivated and mathematically rigorous approach for addressing this challenge. Under DP, a randomized algorithm provides privacy guarantees by approximating the desired functionality, leading to a privacy-utility trade-off. Strong (pure DP) privacy guarantees are often costly in terms of utility. Motivated by the need for a more efficient mechanism with better privacy-utility trade-off, we propose Gaussian FM, an improvement to the functional mechanism (FM) that offers higher utility at the expense of a weakened (approximate) DP guarantee. We analytically show that the proposed Gaussian FM algorithm can offer orders of magnitude smaller noise compared to the existing FM algorithms. We further extend our Gaussian FM algorithm to decentralized-data settings by incorporating the CAPE protocol and propose capeFM. Our method can offer the same level of utility as its centralized counterparts for a range of parameter choices. We empirically show that our proposed algorithms outperform existing state-of-the-art approaches on synthetic and real datasets.

An Examination of Barriers to Accessing Mental Health Care, and Their Association with Depression, Stress, Suicidal Ideation, and Wellness in a Bangladeshi University Student Sample.

BACKGROUND: The mental health burden is high and rising among Bangladeshi university students. Understanding barriers to mental healthcare and how barriers impact mental health outcomes may inform the development of targeted interventions to decrease barriers and improve access to care. AIMS: This study identifies barriers to mental healthcare and their association with mental health outcomes in a Bangladeshi university student sample. METHODS: We conducted a cross-sectional survey (n = 350) on stigma-related, attitudinal, and instrumental barriers to accessing mental healthcare among Bangladeshi university students. We examined the association between stigma and non-stigma (i.e., attitudinal and instrumental) barriers with four mental health outcomes: suicidal ideation, depression, high perceived stress, and wellness. RESULTS: Attitudinal barriers were the most reported barriers. Stigma-related barriers were significant for individuals who had experienced suicidal ideation (aOR = 2.97, p = 0.001), not for individuals with depression. Non-stigma-related barriers were significant for individuals who had experienced depression (aOR = 2.80, p = 0.011). CONCLUSIONS: The current work advances our understanding of how to improve access to mental healthcare among university students in Bangladesh. Stigma-related barriers were particularly salient for individuals who experienced suicidal ideation. Further study is needed on how stigma may impact access to care distinctly for different mental health problems among Bangladeshi university students.

A qualitative exploration of university student perspectives on mindfulness-based stress reduction exercises via smartphone app in Bangladesh.

PURPOSE: Mental health problems are proliferating, and access to mental health care is difficult due to barriers imposed by the COVID-19 pandemic in low-income countries such as Bangladesh. University students are susceptible to mental health concerns, given their unique stressors (i.e., academic pressure, new social environment). Mindfulness techniques can promote mental health , yet their acceptability has not been examined among Bangladeshi university students. These techniques can be used on a digital app, to decrease barriers to use.Qualitative methods were used to examine the acceptability of mindfulness among university students in Bangladesh. In-depth interviews (n = 12) were conducted to examine student reactions to linguistically (Bangla) and culturally adapted mindfulness exercises. Thematic analysis generated three themes (1) previous experience with mindfulness (2) positive responses to and (3) improvements to mindfulness exercises. RESULTS: The results showed favourable attitudes towards the mindfulness content; students expressed positive psychological and physiological reactions. Students welcomed the concept of using these exercises on an app and felt it could overcomepast barriers to help-seeking. CONCLUSIONS: This evidence suggests the value of exploring the acceptability of an app with mindfulness exercises for mental health promotion through a larger-scale pilot study in university students in Bangladesh.

Motivations Toward Using Digital Health and Exploring the Possibility of Using Digital Health for Mental Health in Bangladesh University Students: Cross-sectional Questionnaire Study.

BACKGROUND: Digital health is efficacious for the management and prevention of mental health (MH) problems. It is particularly helpful for the young adult population, who appreciate the autonomy digital health provides, and in low-income countries, where the prevalence of MH problems is high but the supply of professionals trained in MH is low. OBJECTIVE: The objectives of this study are 2-fold: to determine whether university students in Bangladesh find using digital health for MH promotion acceptable and to examine motivational factors for using digital health for MH. METHODS: This study used a cross-sectional survey to examine the likelihood that university students in Bangladesh (n=311) would use different forms of digital health platforms for MH promotion and assessed drivers of intention to use and actual use of digital health generally and digital health for MH through the lens of the Technology Acceptance Model. The results provided evidence that the university student population in Bangladesh is likely to use digital health to promote their MH. RESULTS: Social influence (adjusted odds ratio [aOR] 1.68, 95% CI 1.40-2.01; P<.001), ease of use (aOR 1.85, 95% CI 1.35-2.53; P<.001), and perceived usefulness (aOR 4.12, 95% CI 1.79-9.51; P=.001) of digital health were found to be significant drivers of the intention to use general digital health, and having an intention to use digital health (aOR 2.10, 95% CI 1.17-3.78; P=.01) had the greatest influence on actual use of digital health. Social influence (aOR 1.71, 95% CI 1.43-2.04; P<.001), perceived usefulness (aOR 8.92, 95% CI 4.18-19.04; P<.001), and use of general digital health (aOR 2.16, 95% CI 1.18-3.97; P=.01) were associated with higher intention to use digital health for MH. The use of general digital health (aOR 4.19, 95% CI 2.37-7.41; P<.001) was associated with the actual use of digital health for MH, as were greater non-stigma-related barriers to using traditional clinical MH services (aOR 2.05, 95% CI 1.10-3.80; P=.02). CONCLUSIONS: Overall, we see that the use of digital health for MH is acceptable in this population and can be helpful for students who perceive barriers to receiving traditional care. We also gain insight into how to promote the intention to use digital health, which in turn promotes the actual use of digital health.

Motivations and barriers for clinical mental health help-seeking in Bangladeshi university students: a cross-sectional study.

Background: University and college students are vulnerable to developing depressive symptoms. People in low-income countries are disproportionately impacted by mental health problems, yet few studies examine routes to accessing clinical services. Examining motivation and barriers toward seeking clinical mental health services in university students in Bangladesh is important. Method: Using a cross-sectional survey (n = 350), we assess the relationship between the constructs of autonomy, relatedness, and competency toward using clinical mental health practices (i.e. using professional resources, taking medication) with (1) positive views, (2) perceived need, and (3) use of clinical mental health services among Bangladeshi university students. Results: Results showed that the perceived need for mental health support was the predictor of the largest magnitude (aOR = 4.99, p = 0.005) for using clinical services. Having a positive view of clinical services was predictive of clinical service use (aOR = 2.87, p = 0.033); however, that association became insignificant (p = 0.054) when adjusting for the perceived need for mental health care. Of the SDT constructs, social influences were predictive of perceiving a need for mental health support, and mental health knowledge was predictive (aOR = 1.10, p = 0.001) of having a positive view of clinical mental health care. Conclusion: Our findings show that knowledge of mental health is associated with positive views of mental health services, and that higher levels of stress and the presence of people with mental health problems are associated with the perception of a need for mental health care, which is ultimately responsible for using the services.