Artificial Intelligence for Robotics Assignment 2 Machine Learning
Classify Sonar Targets for Mobile Robot
Assignment 2
Classify Sonar Targets for Mobile Robot
Artificial Intelligence for Robotics
Dr. Min-Fan Ricky Lee
!1
Artificial Intelligence for Robotics Assignment 2 Machine Learning
Classify Sonar Targets for Mobile Robot
1 Summary
Modifying HeartSoundClassification_LiveScript.mlx and conduct the Four Phases of typical machine
learning workflow as shown in Figure 1 to Classify Sonar Targets for Mobile Robot according to Lecture
Slide (“Mobile Learning“) in “Code” Section - “Heart Sound Classifier”
1. Access and Explore Data
2. Preprocess Data
3. Develop Predictive Models
4. Integrate Models with Systems
Figure 1. Workflow: Four Phases of typical machine
2 Task
2.1 Database
Download sonar target dataset [1] as used in the paper [2,3]. Run through the workflow as described
in Summary.
2.2 Sensors
Acquire sonar signal on both the Ground and Surface mobile robot as shown in Figure 2 at T1-102
Autonomous Robotic Laboratory for the classification of sonar returns from two targets, a metal cylinder and a
similarly shaped rock. Run through the workflow as described in Summary.
Figure 2. Mobile robot @T1-102. Left: Ground; Right: Surface
!2
Artificial Intelligence for Robotics Assignment 2 Machine Learning
2.3 Submission
1. Matlab Code
2. Data
3. Word file (EEE Conference Template)
References
[1] Connectionist Bench (Sonar, Mines vs. Rocks) Data Set http://archive.ics.uci.edu/ml/datasets/
connectionist+bench+(sonar,+mines+vs.+rocks)
[2] Gorman, R.Paul ; Sejnowski, Terrence J. Source: Neural Networks, "Analysis of Hidden Units in a Layered
Network Trained to Classify Sonar Targets", v1, n1, p 75-89, 1988
[3] Özkan Bakbak, Pınar, Peker, Musa ," Classification of sonar echo signals in their reduced sparse forms
using complex-valued wavelet neural network, Source: Neural Computing and Applications, 2018
!3