Reading Materials

ML Fundamentals (recommended by Haitham Ammar)

• Gaussian Process for Machine Learning

• Taking the Human Out of the Loop: A Review of Bayesian Optimization

• High-Dimensional Bayesian Optimisation with Variational Autoencoders and Deep Metric Learning

Representation Learning (recommended by Yoshua Bengio)

• Inductive Biases for Deep Learning of Higher-Level Cognition

• Toward Causal Representation Learning

• A Meta-Transfer Objective for Learning to Disentangle Causal Mechanisms

• Learning Neural Causal Models from Unknown Interventions

• Systematic Evaluation of Causal Discovery in Visual Model Based Reinforcement Learning

• Invariance Principle Meets Information Bottleneck for Out-of-Distribution Generalization

Bayesian ML

• Variational Inference: A Review for Statisticians

• Advances in Variational Inference

• Tutorial on Variational Autoencoders

• Auto-Encoding Variational Bayes

Probabilistic causal ML

• Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges

• Geometric foundations of Deep Learning

• Causality: Models, Reasoning, and Inference, Cambridge, 2nd Ed., 2009. Chapters 1-3.

• Elements of Causal Inference: Foundations and Learning Algorithms, MIT Press, 2017. Chapters 1-6.

Gaussian processes (recommended by James Hensman)

• Pattern Recognition and Machine Learning, Introduction to GPs, pages 303 - 320.

• Gaussian Processes for Machine Learning (free download)

Neural models of sequence data, and EHR (recommended by Reza Khorshidi)

• Attention Is All You Need

• BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

• BEHRT: Transformer for Electronic Health Records

• RETAIN: An Interpretable Predictive Model for Healthcare using Reverse Time Attention Mechanism

Computational Pathology (recommended by Lea Goetz)

Computational histopathology reviews:

• Deep neural network models for computational histopathology: A survey

• Deep learning in histopathology: the path to the clinic

Computational histopathology papers covered in Lea Goetz's lecture:

• Clinical-grade computational pathology using weakly supervised deep learning on whole slide images.

• Closing the translation gap: AI applications in digital pathology.

• Development and validation of a deep learning algorithm for improving Gleason scoring of prostate cancer

• Hover-Net: Simultaneous segmentation and classification of nuclei in multi-tissue histology images

• Pathology GAN

• Self-Supervision Closes the Gap Between Weak and Strong Supervision in Histology

• Capturing Cellular Topology in Multi-Gigapixel Pathology Images

• Data-efficient and weakly supervised computational pathology on whole-slide images

Multiple-Instance Learning:

• Multiple-Instance Learning for Medical Image and Video Analysis

• Attention is all you need

• GANs: Generative Adversarial Networks

• Segmentation: U-Net: Convolutional Networks for Biomedical Image Segmentation,

• Contrastive Learning: A Simple Framework for Contrastive Learning of Visual Representations

Causal ML, Interpretability in ML

• AutoML: powering the new human-machine learning ecosystem

• From black boxes to white boxes

• From real-world patient data to individualized treatment effects using machine learning: Current and future methods to address underlying challenges

• Bayesian Inference of Individualized Treatment Effects using Multi-task Gaussian Processes

• Validating Causal Inference Models via Influence Functions

• Estimating Counterfactual Treatment Outcomes over Time through Adversarially Balanced Representations

ML for Climate Action (recommended by David Rolnick)

• "Our World in Data: Emissions by Sector" - context on the sources of GHG emissions

• "Oil in the Cloud" (recommended) - a useful read on some of the negative uses of ML.

• "Tackling Climate Change with Machine Learning" (optional) - David Rolnick's talk will be centred on this paper

Sentiment/Opinion Mining (recommended by Yulan He)

Aspect-Based Sentiment Analysis

• Peng et al. Knowing What, How and Why: A Near Complete Solution for Aspect-Based Sentiment Analysis, AAAI 2020.

• Liang et al. Aspect-invariant Sentiment Features Learning: Adversarial Multi-task Learning for Aspect-based Sentiment Analysis, CIKM 2020.

• Barnes et al., Structured Sentiment Analysis as Dependency Graph Parsing, ACL 2021.

• Cai et al., Aspect-Category-Opinion-Sentiment Quadruple Extraction with Implicit Aspects and Opinions, ACL 2021.

Sentiment-Topic Extraction

• Pergola et al., A Disentangled Adversarial Neural Topic Model for Separating Opinions from Plots in User Reviews, NAACL 2021.

• Zhao et al., Adversarial Learning of Poisson Factorisation Model for Gauging Brand Sentiment in User Reviews, EACL 2021.

Emotion Cause Detection

• Rui et al. RTHN: an RNN-Transformer Hierarchical Network for Emotion Cause Extraction. AAAI 2019.

• Wei et al. Effective Inter-Clause Modeling for End-to-End Emotion-Cause Pair. ACL 2020.

• Yan et al. Position Bias Mitigation: A Knowledge-Aware Graph Model for Emotion Cause Extraction . ACL 2021.

Dialogue Emotion Detection

• Zhong et al., Knowledge-Enriched Transformer for Emotion Detection in Textual Conversations, EMNLP 2019.

• Zhu et al., Topic-Driven and Knowledge-Aware Transformer for Dialogue Emotion Detection, ACL 2021.

Sentiment-Aware Natural Language Generation

• Hu et al. Toward Controlled Generation of Text, ICML 2017.

• Mai et al., Plug and Play Autoencoders for Conditional Text Generation, EMNLP 2020.

• Zhong et al., CARE: Commonsense-Aware Emotional Response Generation with Latent Concepts, AAAI 2021.

Review Question-Answering

• Xu et al. BERT Post-Training for Review Reading Comprehension and Aspect-based Sentiment Analysis. NAACL 2019.

• Deng et al. Opinion-aware Answer Generation for Review-driven Question Answering in E-Commerce. CIKM 2020.

• Lu et al. CHIME: Cross-passage Hierarchical Memory Network for Generative Review Question Answering. COLLING 2020.