Machine learning 101: Where to begin?
summary
Built a camera-controlled feeding closet to recognize which cat may access food and uses that story to walk through data collection, cleaning, augmentation (one image → six variants), two-layer cat feeder with cameras. Uses Python Jupyter + pandas numpy matplotlib scikit-learn and the California housing dataset to show train/validation/test splits (80/10/10 example), scaling, model fitting, and evaluation; shows a raw MSE implying ~$65,000 error, then Random Forest MSE 0.26 versus Gradient Boost MSE 0.29 and discusses overfitting and hyperparameter tuning.
AI-written summary · about
chapters (22)
- 0:00Introduction and welcome
- 1:46About the speaker
- 2:30What is machine learning
- 4:20Types of machine learning
- 6:28Types of data in ML
- 7:12Preparing for machine learning
- 9:03Setting up the environment
- 11:16Data collection and preprocessing
- 14:07Loading and exploring data
- 16:18Cleaning and transforming data
- 18:08Splitting data for training
- 19:42Data augmentation
- 20:26Training and validation sets
- 23:28Top three ML algorithms
- 26:52Training the model
- 27:35Evaluating model performance
- 29:47Confusion matrix explained
- 31:11Mean squared error
- 34:02Improving the model
- 36:37Hyperparameter tuning
- 37:41Real-life ML examples
- 38:46Conclusion and summary
related talks
35mIs it Food? An Introduction to Machine Learning
50mBuilding Intelligent Mobile Apps using Machine Learning
51mMachine Learning Exposed Introduction to Machine Learning
32mYou Can Do Deep Learning!
50mBuilding Intelligent Mobile Apps using Machine Learning
01h 00mA gentle introduction to Deep Learning
31mPut Your Machine Learning on Autopilot
59mA developers guide to Machine Learning