Feature Engineering: Unlocking Potential for Predictive Models
summary
Covers feature engineering techniques to maximize the predictive power of dataset features, improving model flexibility, simplicity, and accuracy. Presents concrete Python examples using Pandas, Scikit-learn, Spark SQL, and TensorFlow to demonstrate these methods.
AI-written summary · about
chapters (21)
- 0:00Introduction
- 1:12Data preparation importance
- 2:54What are features
- 6:55ML competition example
- 9:22Exploratory data analysis
- 11:28ML data structure
- 13:14Raw data transformation
- 16:23Numerical feature techniques
- 22:58Normalization and standardization
- 24:24Feature combination
- 25:49Categorical feature handling
- 27:57Hashing and dimensionality
- 28:40Derived and statistical features
- 29:44Temporal and contextual features
- 30:48Spatial and geographic features
- 32:11Text to features
- 32:53Text modeling topics
- 33:55Feature selection
- 34:56Features impact on models
- 35:39Automation in feature engineering
- 35:59References and conclusion
related talks
02h 17mFeature Engineering is for Everyone!
30mFeature Engineering: An Apprentice’s Guide to the “Dark Art” of Machine Learning
01h 28mFeature Engineering Made Simple
35mTime-series Feature Engineering Done Right
48mThe Missing Step in Moving from Data to Models
01h 25mUnlocking the Predictive Power of Relational Data with Automated Feature Engineering
43mFeature Engineering for Time Series Forecasting
31mLife cycle of data becoming a model