Machine Learning Street Talk (MLST) Sayak Paul
Jul 17, 2020
Sayak Paul, a prominent figure in deep learning and Google Developer Expert, shares insights from his vibrant career in machine learning. He discusses the AI landscape in India and the nuances of unsupervised representation learning. The conversation dives into data augmentation and contrastive learning techniques, emphasizing their importance in performance improvement. Sayak further explores the complexities of explainability and interpretability in AI, suggesting ethical responsibilities for engineers. The talk wraps up with advanced topics on pruning and the lottery ticket hypothesis in neural networks.
AI Snips
Chapters
Books
Transcript
Episode notes
Learning Through Teaching and Reasoning
- Understand machine learning concepts by teaching and discussing them with others.
- Motivate your learning by understanding the reasons behind a concept's development, such as residual networks.
Balancing Practical Implementation with Research
- Focus on a narrow approach to research papers; read abstracts first, then watch explanatory videos if interested.
- Implement concepts by creating minimal code examples if you feel excited after understanding the paper.
Colab Productivity Tip
- Sayak Paul uses Google Colab for about two years and finds it productive.
- He recommends storing data in Google Cloud Storage buckets for faster access, especially with TPUs.
