Abstract: Graph Learning

Graphs represent a fundamental shift in data analysis compared to conventional tabular-based machine learning. This talk provides an in-depth exploration of "Graph Learning" and its distinctions from traditional data analysis methods. Beginning with the foundational concepts of graphs, we delve into the intricacies of nodes and edges, emphasizing their pivotal role in capturing complex relationships and structures. Graphs offer a unique lens through which we can analyze data, revealing insights not easily attainable through traditional means. However, the focus of this presentation extends beyond theory to practical applications within the enterprise landscape. Knowledge graphs, a type of graph database, have revolutionized data management for organizations. Discover how knowledge graphs can transform data silos into interconnected knowledge networks, enabling enhanced decision-making, recommendation systems, and semantic search capabilities. The discussion also delves into specialized algorithms and techniques tailored for knowledge graphs. These methods play a crucial role in extracting valuable information from vast interconnected datasets, aiding in the identification of patterns, anomaly detection, and the optimization of data-driven strategies across various industries.

Abstract : Generative AI

Tutorial session provides an introduction to Generative AI and explores the role of large language models in the industry, both in the present and future. The session also includes hands-on demonstrations of generative AI in computer vision, focusing on DALLE and Stable Diffusion. Additionally, participants will have the opportunity to experience hands-on activities with large language models, such as GPT-4 and PaLM-E. Join us to delve into the exciting world of Generative AI and discover the potential it holds for various applications.

Prerequisites:  

• Basic Knowledge of AI, ML and DL.

• Sound Knowledge of Python programming and Jupyter/Colab Notebook

Abstract : Foundation Models

• The tutorial will begin with an exploration of how foundation models have revolutionized NLP, enabling machines to understand, generate, and manipulate human language like never before. We will uncover the architecture and training techniques behind these models, discussing their underlying mechanisms that enable tasks such as language translation, sentiment analysis, and question answering.

• Transitioning into the realm of Computer Vision, we will explore how foundation models have reshaped image recognition, object detection, and even artistic style transfer. We will unravel the methods that allow these models to extract intricate details from images, making them capable of classifying objects, generating captions, and even assisting in medical diagnoses.

• The workshop will then journey into the world of Audio, where foundation models have found applications in speech recognition, music generation, and sound classification. Participants will gain insights into the neural architectures that empower these models to transcribe spoken words, compose melodies, and identify environmental sounds.