Despite the high research activity in robustness machine learning, there is little exploration of adaptive algorithms that can maintain performance in dynamically changing environments. Additionally, there is a gap in developing transparent methodologies for assessing and quantifying robustness in real-world applications, particularly in safety-critical systems.
### 2.
THE OPPORTUNITY
The high volume of 438 publications coupled with only 28 open roles suggests that while the field is burgeoning, there may be a limited number of opportunities relative to the amount of new research being produced. This indicates that a researcher entering this space now has the chance to distinguish themselves by focusing on underserved areas, potentially leading to niche expertise that could be highly sought after.
### 3.
Explore the development of hybrid models that combine classical statistical methods with modern machine learning techniques to enhance robustness in unstable environments.
Search papers →
Investigate the role of explainable AI in assessing the robustness of machine learning models through intuitive visualizations and enhanced user interpretability.
Search papers →
Study the implications of adversarial attacks on offline versus online learning systems, particularly focusing on how different learning paradigms affect model resilience.
Search papers →
+ abstract
Chromatin organisation influences gene regulation and genome stability, yet its dysregulation in cancer remains incompletely understood. This has become particularly important for patient diagnostics using cell-free DNA (cfDNA) from body fluids, based on computational analyses of nucleosome occupancy landscapes reconstructed from cfDNA. This thesis establishes the first comprehensive atlas of nucl...
+ abstract
Online neural learning requires models that update after each incoming example, remain calibrated under distributional change, and avoid brittle gradient transmission. The original version of this work used a small static benchmark, a shallow model, few random seeds, and no significance tests. This revision reformulates the problem as genuine prequential online learning and introduces Smooth Onlin...
+ abstract
Do large language models (LLMs) possess a measurable "personality," and how do the linguistic properties of training corpora shape their cognitive style and downstream reasoning? This paper approaches these questions from a sociolinguistic perspective on machine "identity." This manuscript is positioned explicitly as a conceptual perspective paper: it does not present original experimental data, n...
+ abstract
As generative artificial intelligence and automated content curation rapidly reshape the global media landscape, the intersection of algorithmic justice and media governance has become a critical frontier for sustainable development. This study provides a comparative communication policy analysis of China and South Korea, focusing explicitly on how their distinct regulatory toolkits address the te...
+ abstract
Continuous Emission Monitoring Systems (CEMS) for stationary pollution sources serve as the core data backbone for precision pollution control and environmental law enforcement. However, traditional data quality control (QC) primarily relies on fixed-threshold screening and manual spot checks, making it highly challenging to effectively identify sophisticated anomalies and possible artificial inte...
+ abstract
As artificial intelligence gets deeply embedded into immersive environments, attention-guiding technology delivers ever more precise and efficient outcomes. Still, this improved efficiency carries an overlooked design risk: excessive AI guidance may steadily erode user agency and undermine exploratory behaviors as well as the process of meaning-making. This paper builds a critical conceptual frame...
+ abstract
Segregation research increasingly measures who people encounter beyond home, yet physical mobility, place semantics, and digital activity are often combined without clear construct boundaries. This study develops a physical--digital activity-space framework that distinguishes co-presence, digital exposure, and digital interaction, and represents physical and digital segregation as separate compone...
+ abstract
+ abstract
+ abstract
Uncertainty propagation is essential for reliable engineering design, but when input distributions are unknown, or data are scarce, probabilistic models risk misspecification, making interval representations a more robust alternative. However, interval propagation requires solving optimisation problems for each bound, leading to high computational cost, especially in downstream tasks such as desig...