Deep Learning for Time Series Made Easy
summary
mcfly is an open-source Python library (a Keras wrapper) that helps non-experts apply deep learning to time series classification by providing ready-made architectures and an automatic hyperparameter search. The talk shows using mcfly on multi-channel sensor time series (wrist accelerometer / movement data) to train models that classify activity types such as sleeping, walking and climbing stairs.
AI-written summary · about
chapters (13)
- 0:00Deep learning for time series
- 1:10Why deep learning
- 3:38Time series projects
- 5:02Advantages over traditional
- 7:08Neural network basics
- 8:56Convolutional layers
- 11:03Recurrent layers
- 13:29Choosing architectures
- 14:35Hyperparameter optimization
- 16:01Introducing McFly
- 18:28Training models
- 20:14Model performance
- 21:38Conclusion
related talks
26mTime Series Analysis in Python: Easy Tools for Scientific Insight
48m(Deep) Learning to Fly
42mFlying Python at 40,000 feet
30msktime: python toolbox for time series- next-generation AI
01h 02mDeep learning for time series forecasting and classification in practice
31mAdvanced machine learning time series....
01h 27mIntroduction to Machine Learning with Time Series
01h 00m(Deep) Learning to Fly