3.5 Machine learning in gait recognition
A person's movement is characterized by arbitrary. Machinelearning
technology can be adopted to get characters, which
is customized for different user and different situation. The
different stage of a gait can be recognized from machine
learning techniques. The mainly data classification processes
are as follows:
1) Acquisition of sensor data;
The process of machine learning is based on large amounts of
data. The raw data can be obtained by sensors, transmitted by
CAN bus, and then gathered to work station through the
Ethernet port.
2) Unsupervised analysis of data;
In unsupervised learning, clustering algorithm is used to
classify sample data based on the similarity among sample
data. The mathematical definition is as follows: