MLx Health & Bio 25

This module delves into the intricacies of survival analysis and causal inference within the healthcare sector. Students will learn to navigate biases in medical data, apply advanced statistical methods, and integrate machine learning techniques to enhance decision-making in clinical settings. The knowledge gained will empower students to critically assess and improve health-related research methodologies.

1. 1. ML for data with left and right censorships
2. 2. Q&A
3. 3. Algorithms for decision making; blending DL with causal inference
4. 4. Q&A

This module delves into the transformative role of AI in biomedicine, emphasizing precision health and the integration of multimodal data. Students will explore the challenges of real-world data, including disease imbalance and the use of patient embeddings, while learning about innovative solutions for clinical trials and diagnosis accuracy. This knowledge is crucial for advancing healthcare accessibility and improving patient outcomes.

1. 1. Talk
2. 2. Q&A

This module delves into the critical evaluation of social biases in AI-generated text, focusing on gender and ethnic disparities. Students will learn to develop and apply bias score algorithms to analyze job-related content, understand the implications of these biases in the labor market, and explore strategies for mitigating their effects. This knowledge is essential for fostering diversity and equity in AI applications.

1. 1. Talk

This module delves into the transformative role of generative AI in healthcare, covering the architecture of transformer models, prompt engineering, and the development of specialized medical models. Students will gain insights into the advantages and challenges of AI in clinical settings, while also addressing ethical considerations and the necessity for domain-specific knowledge.

1. 1. Talk

This module delves into the integration of wearable devices and machine learning within population health research. Students will explore the significance of time series data, the impact of large-scale health studies, and innovative self-supervised learning techniques. By understanding these concepts, learners will be equipped to leverage technology for improved health monitoring and data analysis, paving the way for future advancements in healthcare.

1. 1. Talk
2. 2. Q&A

This module, "ML for Environmental Epidemiology and the Study of Exposome," delves into the intersection of machine learning and environmental health. It begins with an overview of environmental epidemiology, emphasizing the effects of air pollution and temperature on health, while addressing traditional statistical methods and the challenges of exposure assessment. The second unit focuses on linear mixed effect models, exploring estimation techniques, convergence issues, and the integration of machine learning to enhance modeling. Participants will gain insights into the complexities of environmental factors affecting health outcomes and the innovative approaches to improve research methodologies.

1. 1. Talk
2. 2. Q&A

Causal Representation Learning for Biomedical Applications delves into the intricacies of causal effect estimation, emphasizing the significance of interventions and confounders in biomedical contexts. The module covers foundational concepts such as average treatment effects and instrumental variable regression, while also addressing the critical role of context in causal inference. Advanced statistical methods, including sensitivity analysis and the integration of causal inference with reinforcement learning, are explored, alongside challenges posed by high-dimensional data and the importance of regularization techniques. Participants will gain practical insights into proxy methods and uncertainty estimation in causal effect analysis.

1. 1. Talk
2. 2. Q&A

This module, "On-Device Deployment of ML in Health," delves into the integration of on-device AI technologies within healthcare settings, focusing on privacy, real-time processing, and cost efficiency. Participants will learn about the pivotal role of PyTorch in AI model development, the challenges of deploying these models on edge devices, and the innovative solution offered by Executorch for efficient deployment. The module also covers various applications of on-device AI, such as diagnostics and health monitoring, while addressing technical innovations and optimization techniques. Additionally, it explores Executorch's capabilities for embedded devices, including continual learning and model segmentation for resource-constrained environments.

1. 1. Talk
2. 2. Q&A

This module delves into the integration of machine learning with biological data, focusing on DNA and protein sequences. Students will learn about the central dogma of molecular biology, genome language modeling, and the significance of tokenization strategies for variant effect prediction. By exploring the relationship between genotype and phenotype, participants will gain insights into designing innovative biological systems using AI.

1. 1. Talk
2. 2. Q&A

This module delves into the transformative role of AI in scientific and medical research, focusing on projects that enhance hypothesis generation, clinical diagnostics, and multimodal data processing. Students will learn how AI co-scientists can amplify research capabilities, improve patient interactions, and address real-world challenges in healthcare delivery.

1. 1. AI Co-scientist
2. 2. Q&A1
3. 3. AI co-physician
4. 4. Q&A2

This module delves into the integration of AI and machine learning in infectious disease modeling, addressing challenges like data fragmentation and bias. Students will learn about time series forecasting, the impact of societal factors on disease spread, and the importance of genomic data. By exploring privacy-enhancing techniques in federated learning, participants will gain insights into effective public health interventions and decision-making.

1. 1. Talk
2. 2. Q&A

This module delves into the intersection of geometric deep learning and its applications in health and bioinformatics. Students will explore the significance of symmetry, equivariance, and invariance in neural networks, and how these concepts enhance model performance in drug design and computational biology. Additionally, the module addresses the challenges of hardware efficiency and model generalization, equipping students with a comprehensive understanding of advanced machine learning architectures.

1. 1. Talk
2. 2. Q&A

This module delves into the integration of multimodal and multiomics data in cancer research and clinical settings. Students will learn methodologies for data fusion, the significance of diverse data types, and advanced modeling techniques to enhance predictive accuracy in therapy response and cancer detection. The knowledge gained will empower students to tackle real-world challenges in data analysis and improve clinical outcomes.

1. 1. Talk
2. 2. Q&A

This module delves into the transformative role of on-device AI in healthcare, emphasizing local processing benefits such as privacy and low latency. Students will learn about the Lama.cpp project for efficient AI model deployment, explore financial implications of on-device solutions, and master model fine-tuning techniques to optimize performance on embedded devices.

1. 1. Talk
2. 2. Q&A

This module delves into the development and application of vision language models (VLMs), emphasizing prompt learning and Bayesian methods. Students will explore how these models integrate visual and textual data, improve performance through advanced techniques, and address challenges in multimodal applications. The knowledge gained will be crucial for leveraging AI in diverse domains.

1. 1. Talk
2. 2. Q&A

This module delves into the transformative role of AI in medical imaging and neural disease detection. Students will explore advanced techniques such as segmentation, autoencoders, and the integration of fMRI and EEG for enhanced diagnostic capabilities. Emphasis is placed on ethical considerations, model interpretability, and the challenges of explainability in clinical AI applications.

1. 1. Talk
2. 2. Q&A