Behavior Analysis with Machine Learning and R: The free eBook


By Enrique Garcia-Ceja, Researcher at SINTEFdigital, R + habits evaluation + sensors + machine learning.

Automatic habits monitoring applied sciences have gotten a part of our on a regular basis lives due to advances in sensors and machine learning. The automated evaluation and understanding of habits are being utilized to resolve issues in a number of fields, together with well being care, sports activities, advertising, ecology, safety, and psychology, to call a number of. Behavior Analysis with Machine Learning and R goals to supply an introduction to machine learning ideas and algorithms utilized to a various set of habits evaluation issues. It focuses on the sensible features of fixing such issues based mostly on knowledge collected from sensors or saved in databases.

The guide covers subjects inside the complete knowledge evaluation pipeline—from knowledge assortment, visualization, preprocessing, and encoding to mannequin coaching and analysis. No prior information in machine learning is assumed. Some of the subjects embody:

  • How to construct supervised studying fashions to foretell indoor areas based mostly on Wi-Fi indicators, acknowledge bodily actions from smartphone sensors and 3D skeleton knowledge, detect hand gestures from accelerometer indicators, and rather more.
  • Learn how unsupervised studying algorithms can be utilized to find prison behavioral patterns and learn how Miss Karlene used affiliation guidelines mining to assist this poor woman discover her stolen doll:

  • Program your personal ensemble studying strategies and use multi-view stacking to fuse indicators from heterogeneous knowledge sources.
  • Train deep studying fashions with Keras and TensorFlow, together with neural networks to categorise muscle exercise from electromyography indicators and convolutional neural networks to detect smiles in photographs.

The guide is accessible for free at


  • Chapter 1. Introduction
  • Chapter 2. Predicting Behavior with Classification Models
  • Chapter 3. Predicting Behavior with Ensemble Learning
  • Chapter 4. Exploring and Visualizing Behavioral Data
  • Chapter 5. Preprocessing Behavioral Data
  • Chapter 6. Discovering Behaviors with Unsupervised Learning
  • Chapter 7. Encoding Behavioral Data
  • Chapter 8. Predicting Behavior with Deep Learning
  • Chapter 9. Multi-User Validation
  • Appendix A: Setup your Environment
  • Appendix B: Datasets


Bio: Enrique Garcia Ceja is a analysis scientist at SINTEF, Norway. For the final 10 years, he has been conducting analysis on habits monitoring and evaluation with machine learning and wearable units.



Source hyperlink

Write a comment