Biological knowledge is widely represented in the form of ontologies and ontology-based annotations. Biomedical ontologies describe known phenomena in biology using formal axioms, and the annotations associate an entity (e.g. genes, diseases, chemicals, etc.) with a set of biological concepts. In addition to formally structured axioms, ontologies contain meta-data in the form of annotation properties expressed mostly in natural language which provide valuable pieces of information that characterize ontology concepts. The structure and information contained in ontologies and their annotations make them valuable for use in machine learning, data analysis and knowledge extraction tasks. I develop the first approaches that can exploit all of the information encoded in ontologies, both formal and informal, to learn feature embeddings of biological concepts and biological entities based on their annotations to ontologies. Notably, I develop the first approach to use all the formal content of ontologies in the form of logical axioms and entity annotations to generate feature vectors of biological entities using neural language models. I extend the proposed algorithm by enriching the obtained feature vectors through representing the natural language annotation properties within the ontology meta-data as axioms. Transfer learning is then applied to learn from the biomedical literature and apply on the formal knowledge of ontologies. To optimize learning that combines the formal content of biomedical ontologies and natural language data such as the literature, I also propose a new approach that uses self-normalization with a deep Siamese neural network that improves learning from both the formal knowledge within ontologies and textual data. I validate the proposed algorithms by applying them to the Gene Ontology to generate feature vectors of proteins based on their functions, and to the PhenomeNet ontology to generate features of genes and diseases based on the phenotypes they are associated with. The generated features are then used to train a variety of machinelearning based classifiers to perform different prediction tasks including the prediction of protein interactions, gene–disease associations and the toxicological effects of chemicals. I also use the proposed methods to conduct the first quantitative evaluation of the quality of the axioms and meta-data included in ontologies to prove that including axioms as background improves ontology-based prediction. The proposed approaches can be applied to a wide range of other bioinformatics research problems including similarity-based prediction and classification of interaction types using supervised learning, or clustering.

Biological knowledge is widely represented in the form of ontologies and ontologybased annotations. Biomedical ontologies describe known phenomena in biology using formal axioms, and the annotations associate an entity (e.g. genes, diseases, chemicals, etc.) with a set of biological concepts. In addition to formally structured axioms, ontologies contain meta-data in the form of annotation properties expressed mostly in natural language which provide valuable pieces of information that characterize ontology concepts. The structure and information contained in ontologies and their annotations make them valuable for use in machine learning, data analysis and knowledge extraction tasks. I develop the rst approaches that can exploit all of the information encoded in ontologies, both formal and informal, to learn feature embeddings of biological concepts and biological entities based on their annotations to ontologies. Notably, I develop the rst approach to use all the formal content of ontologies in the form of logical axioms and entity annotations to generate feature vectors of biological entities using neural language models. I extend the proposed algorithm by enriching the obtained feature vectors through representing the natural language annotation properties within the ontology meta-data as axioms. Transfer learning is then applied to learn from the biomedical literature and apply on the formal knowledge of ontologies. To optimize learning that combines the formal content of biomedical ontologies and natural language data such as the literature, I also propose a new approach that uses self-normalization with a deep Siamese neural network that improves learning from both the formal knowledge within ontologies and textual data. I validate the proposed algorithms by applying them to the Gene Ontology to generate feature vectors of proteins based on their functions, and to the PhenomeNet ontology to generate features of genes and diseases based on the phenotypes they are associated with. The generated features are then used to train a variety of machinelearning based classi ers to perform di erent prediction tasks including the prediction of protein interactions, gene{disease associations and the toxicological e ects of chemicals. I also use the proposed methods to conduct the rst quantitative evaluation of the quality of the axioms and meta-data included in ontologies to prove that including axioms as background improves ontology-based prediction. The proposed approaches can be applied to a wide range of other bioinformatics research problems including similarity-based prediction and classi cation of interaction types using supervised learning, or clustering.

In the past decade, reinforcement learning (RL) has achieved breakthroughs across various domains, from surpassing human performance in strategy games to enhancing the training of large language models (LLMs) with human feedback. However, RL has yet to gain widespread adoption in mission-critical fields such as healthcare and autonomous vehicles. This is primarily attributed to the inherent lack of trust, explainability, and generalizability of neural networks in deep reinforcement learning (DRL) agents. While neural DRL agents leverage the power of neural networks to solve specific tasks robustly and efficiently, this often comes at the cost of explainability and generalizability. In contrast, pure symbolic agents maintain explainability and trust but often underperform in high-dimensional data. In this work, we developed a method to distill explainable and trustworthy agents using neuro-symbolic AI. Neuro-symbolic distillation combines the strengths of symbolic reasoning and neural networks, creating a hybrid framework that leverages the structured knowledge representation of symbolic systems alongside the learning capabilities of neural networks. The key steps of neuro-symbolic distillation involve training traditional DRL agents, followed by extracting, selecting, and distilling their learned policies into symbolic forms using symbolic regression and tree-based models. These symbolic representations are then employed instead of the neural agents to make interpretable decisions with comparable accuracy. The approach is validated through experiments on Lunar Lander and Pong, demonstrating that symbolic representations can effectively replace neural agents while enhancing transparency and trustworthiness. Our findings suggest that this approach mitigates the black-box nature of neural networks, providing a pathway toward more transparent and trustworthy AI systems. The implications of this research are significant for fields requiring both high performance and explainability, such as autonomous systems, healthcare, and financial modeling.

Modern autonomous systems should be able to interact with their surroundings in a flexible yet safe manner. To guarantee this behavior, such systems must learn how to approach unseen entities in their environment through the inference of relationships between actions and objects, called affordances. This research project introduces a neuro-symbolic AI system capable of inferring affordances using attribute detection and knowledge representation as its core principles. The attribute detection module employs a visuo-lingual image captioning model to extract the key object attributes of a scene, while the cognitive knowledge module infers the affordances of those attributes using conditional probability. The practical capabilities of the neuro-symbolic AI system are assessed by implementing a simulated robot system that interacts within the problem space of jars and bottles. The neuro-symbolic AI system is evaluated through its caption-inferring capabilities using image captioning and machine translation metrics. The scores registered in the evaluation show a successful attribute captioning rate of more than 71%. The robot simulation is evaluated within a Unity virtual environment by interacting with 50 jars and bottles, equally divided between lifting and twisting affordances. The robot system successfully interacts with all the objects in the scene due to the robustness of the architecture but fails in the inference process 24 out of the 50 iterations. Contrary to previous works approaching the problem as a classification task, this study shows that affordance inference can be successfully implemented using a cognitive visuo-lingual method. The study’s results justify further study into the use of neuro-symbolic AI approaches to affordance inference.