AIAI Seminar-10 June-Talk by Lauren DeLong and Nijesh Upreti

Speaker: Lauren DeLong

Title: Investigating associations between physical multimorbidity and subsequent depression via a systematic cluster analysis

Abstract: Multimorbidity, the co-occurrence of two or more conditions within an individual, is a growing challenge for health and care delivery as well as for research Many multimorbidity studies focus upon the co-occurrence of physical health conditions, but mental health disorders are less represented. However, recent studies have revealed links of a bidirectional nature between depression and physical conditions. To investigate associations between physical multimorbidity and subsequent depression, we first performed a clustering analysis upon baseline morbidity data for UK Biobank participants. In contrast to previous similar studies, we compared the usefulness of four independent clustering methods. The identified clusters indicated which physical conditions tend to co-occur most frequently in the whole population and stratified by sex. Finally, we used survival analysis to compare time to subsequent depression diagnosis between participants with particular groups of physical conditions at baseline and those without physical conditions at baseline. In comparison to agglomerative hierarchical clustering, latent class analysis, and k¬-medoids, we found that k-modes models showed the best clustering performance amongst several metrics. Notably, the differentially represented conditions within several clusters reflected known bodily systems, such as the respiratory or digestive systems. While we found that certain clusters had stronger associations with depression, we also noted a positive correlation between such associations and the average number of conditions per participant. Therefore, both the type and number of conditions likely influence the subsequent diagnosis of depression. Our findings suggest further investigation into other factors, like social ones, which may link the effects of physical multimorbidity and depression.

Speaker: Nijesh Upreti

Title: Towards Logical Characterisation of Neuro-symbolic Weak Supervision

Abstract: We address the challenges of multi-instance weak supervision by segmenting the problem space using auxiliary predicates and applying Inductive Logic Programming (ILP) to provide a logical characterization for the domain. This characterization ensures that learning tasks are well-structured and verifiable. Specifically, we investigate the learnability of the Transition Predicate (TP) when the Classifier Predicate (CP) and Observed Predicate (OP) are known, and conversely, we explore how CP can be inferred given known TP and OP. This bidirectional methodology enables comprehensive analysis and validation of both the transition functions and the efficacy of neural network classifiers within a clearly defined weak supervision context. By partitioning the problem into manageable components via auxiliary predicates, we lay a logical groundwork that enhances both the interpretability and robustness of the learning models. This approach not only clarifies the semantics of weak supervision but also facilitates the verification of the models against predefined logical standards. Our strategy extends the conventional framework of weak supervision by integrating neuro-symbolic elements, thus bridging the gap between symbolic logic and neural computation, and expanding the potential of machine learning in environments where traditional supervision is inadequate.