The Evolution of Supervised Machine Learning Approaches for Text Analysis
summary
Evolution of supervised machine learning approaches for text analysis, covering classical methods (feature engineering, bag-of-words, linear models) and newer techniques (word embeddings and neural networks). Discusses practical issues like dataset creation, label quality, evaluation metrics, and how model choices evolved to address these problems.
AI-written summary · about
chapters (7)
- 0:00Start
- 9:57Logistic regression
- 10:47Case 3 - Commercial filter
- 11:32Classic NLP stack
- 17:35Old-fashioned Neural networks
- 20:14Deep Learning Formula for NLP
- 20:51Recurrent Neural Network
related talks
06mAdvancements in Natural Language Processing (NLP) and Text Analytics
45mActive learning and weak supervision in NLP projects
42mThe Machine Understands Text — Learning and Search at Micro and Macro Scale
41mResearch in the Age of AI
52mMachine learning: from practice to theory
49mUsing Machine Learning and Not Dying Trying
03mGoing beyond supervised learning
02h 11mGunnvant Singh Saini of Jigsaw Academy at Machine Learning Developers Summit 2019