OxML 2020

This module delves into the principles of Bayesian machine learning, emphasizing the quantification of uncertainty in decision-making. Students will learn about probabilistic models, approximate inference methods, and their applications in fields like healthcare and recommender systems. By the end, participants will be equipped to apply advanced techniques such as variational inference and partial variational autoencoders to tackle real-world challenges.

1. 1. Talk

This module delves into the intricacies of Gaussian processes, equipping students with a solid understanding of Gaussian distributions, kernels, and variational inference. Learners will explore practical applications in regression and classification, while mastering hyperparameter optimization through marginal likelihood. This knowledge is essential for developing robust machine learning models.

1. 1. Talk

This module offers a comprehensive introduction to deep learning, emphasizing its pivotal role in computer vision. Students will explore the architecture of deep neural networks, the functionality of convolutional layers, and the intricacies of training models with extensive datasets. Additionally, the module addresses the evolution of deep learning techniques, their limitations, and the significance of understanding model behavior for real-world applications.

1. 1. Talk

This module delves into the evolution and techniques of representation learning in deep learning, emphasizing unsupervised methods. Students will explore the philosophical foundations, differentiate between learning paradigms, and understand the significance of efficient representations for various applications. By examining techniques like autoencoders and contrastive learning, learners will gain insights into the challenges and ongoing research in this dynamic field.

1. 1. Talk

This module delves into the relationship between language representation in the brain and computational models, particularly focusing on neural networks like RNNs and transformers. Students will learn how these models interpret brain activity to predict word meanings and the role of context in language processing, enhancing their understanding of multilingual language representation.

1. 1. Talk

This module delves into the integration of machine learning with electronic health records, highlighting AI's transformative role in healthcare. Students will explore various machine learning models, the challenges of multimorbidity, and the significance of effective data representation in predicting health outcomes and improving patient care.

1. 1. Talk

This module delves into the integration of geometric and topological methods in data science, focusing on their applications in analyzing complex datasets. Students will learn to generate topological models, understand the significance of shape in data analysis, and explore the relationship between topological data analysis (TDA) and deep learning, equipping them with valuable insights for feature extraction in machine learning.

1. 1. Talk

This module delves into the integration of machine learning within the healthcare sector, addressing the specific challenges and opportunities it presents. Students will learn about automated machine learning, the significance of model interpretability, and the necessity of collaboration with healthcare professionals to enhance clinical decision-making and patient outcomes.

1. 1. Talk

This module explores the transformative role of deep learning in healthcare, focusing on medical imaging and the evolution of technology in this field. Students will learn about federated learning for maintaining patient data privacy, and gain hands-on experience with NVIDIA Clara and Monai toolkits, which are essential for implementing AI solutions in medical workflows.

1. 1. Talk

This module delves into the integration of machine learning techniques within medical imaging, focusing on data harmonization, segmentation, and the deployment of deep learning models like UNet. Students will learn about the challenges faced in image registration and the critical role of explainable AI in enhancing medical diagnostics, equipping them with the knowledge to innovate in clinical applications.

1. 1. Talk

This module delves into the fundamentals of reinforcement learning, emphasizing its practical applications in gaming, robotics, and healthcare. Students will learn to navigate the exploration-exploitation trade-off, differentiate between bandit problems and reinforcement learning, and apply key algorithms like Q-learning and SARSA to real-world scenarios. This knowledge is essential for developing adaptive systems that can learn and improve over time.

1. 1. Talk

This module delves into the critical relationship between AI hardware and neural network design, equipping students with the knowledge to optimize machine learning models across various architectures. Key topics include performance modeling, compute intensity, and advanced optimization techniques, providing insights essential for enhancing computational efficiency while maintaining model accuracy in future AI applications.

1. 1. Talk

This module delves into the integration of genomics and machine learning, equipping students with the skills to analyze genomic data through computational methods. Participants will learn to process and interpret data related to population genetics, disease genetics, and functional genomics, while applying machine learning techniques to predict health outcomes and infer evolutionary parameters.

1. 1. Talk

This module delves into the transformative role of machine learning in product and drug discovery, emphasizing the integration of big data and genetic information. Students will explore real-world applications through case studies, understand the infrastructure needed for health tech, and identify challenges in scaling machine learning models. This knowledge is crucial for advancing personalized medicine and therapeutics.

1. 1. Talk