Tutorials

Unveiling Graph Structures for Machine Learning: Learning, Structuring, and Coarsening

Manoj Kumar (IIT Delhi) and Sandeep Kumar (IIT Delhi)

This tutorial covers the key aspects of graph machine learning: learning graphs from data, structuring graphs, and graph dimensionality reduction with its utility in diverse applications.
Synthetic Data Generation for Enterprise DBMS

Anupam Sanghi (IBM Research) and Jayant Haritsa (Indian Institute of Sciene)

This tutorial delves into synthetic data generation for enterprise DBMS testing, focusing on ab initio, database-dependent, and workload-dependent methods. It outlines potential research directions and challenges.
Contrastive learning: Big Data Foundations and Applications

Sandhya Tripathi (Washington University School of Medicine in St Louis) and Christopher Ryan King (Washington University School of Medicine in St Louis)

This tutorial will review foundations and applied practices in contrastive learning in vision, text, time series and tabular data modalities with a complementary code base.
Building Knowledge Graph for Products at Web Scale

Manoj Agarwal (Uber)

Building a Product Knowledge Graph is key to enable intelligent entity/product search in many ways. In this tutorial, we cover various aspects such as: 1) Why the knowledge graph for products is the key to semantic search and product discovery?; 2) The key challenges and methodologies in building such a knowledge graph at scale for a highly diverse product landscape; 3) Open research questions.
Unlocking the Potential of Unstructured Data in Business Documents Through Document Intelligence

Himanshu Sharad Bhatt (American Express), Sriranjani Ramakrishnan (American Express), Sachin Raja (IIIT Hyderabad) and C. V. Jawahar (IIIT Hyderabad)

This tutorial provides a comprehensive overview of intelligent document processing by leveraging unlabelled data with modern deep learning architectures handling unified text, image, layout, style and other information for multiple downstream applications in industry.
Beyond Point Estimates: Theory & Applications of Uncertainty Estimates for Supervised Models

Arunita Das (Amazon), Sanyog Dewani (Amazon), Srujana Merugu (Amazon) and Gokul Swami (Amazon)

This tutorial delves into a) SOTA research on uncertainty estimation algorithms, b) evaluation metrics of uncertainty estimation methods c) applications of uncertainty estimates with a special focus on supervised problems (d) hands on session on building uncertainty estimation models
Deep Tractable Probabilistic Models

Athresh Karanam (The University of Texas at Dallas), Saurabh Mathur (The University of Texas at Dallas), Sahil Sidheekh (The University of Texas at Dallas) and Sriraam Natarajan (The University of Texas at Dallas)

This tutorial provides a unified and comprehensive view of generative models that can (a) learn complex data distributions and (b) perform exact probabilistic inference tractably to reason about the world and make reliable decisions.
Towards understanding and mitigating the hallucinations in NLP and Speech

Ashish Mittal (IBM Research), Rudra Murthy (IBM Research), Vishwajeet Kumar (IBM Research) and Riyaz Bhat (IBM Research)

This tutorial covers the hallucination phenomenon observed in generative AI models like autoregressive models and sequence-to-sequence models. The tutorial also provides a bird's eye-view of the approaches to mitigate hallcuinations.
Counter Turing Test (CT^2): Identifying the Origin of Texts - Human or AI?

Megha Chakraborty (AI Institute, University of South Carolina), Amitava Das (AI Institute, University of South Carolina), Aman Chadha (Stanford University, Amazon Science) and Amit Sheth (AI Institute, University of South Carolina)

This tutorial delves into a comprehensive exploration of the various techniques formulated for AI-Generated Text Detection (AGTD). To date, seven distinct technique categories have been identified, each offering its own merits and grappling with distinct limitations. We offer to provide empirical evidence on why AGTD is an unsolved problem to date.
Tutorial on Fair and Private Deep Learning

Manisha Padala (Indian Institute of Science), Sankarshan Damle (IIIT Hyderabad) and Sujit Gujar (IIIT Hyderabad)

This tutorial provides an exhaustive discussion on fairness and privacy challenges in DL-based classification and existing solutions to address them, both separately and simultaneously.