Human Activity Data Clustering & Classification

This example demonstrates analyzing a data set consisting of derived measurements
from cellphone sensors worn by individuals while performing certain activities
such as walking and standing. The first goal of this example is to understand
the unlabeled measurement dataset through visualization and clustering. The second
goal is to build a predictive model for classifying activity type given labelled
training data.

Look at the readme.txt file in the Data folder for more information on the dataset.

Visualization and Clustering:
The scripts ClusterAnalysis and ClusterAdvanced demonstrate performing clustering
and visualization of the dataset. ClusterAnalysis focuses on K-means clustering 
while ClusterAdvanced compares numerous clustering techniques. In this example
there is a strong emphasis on understanding the data, its structure, features and 
discerning meaning from the clustering analysis.

Predictive Modeling:
The scripts ClassificationAnalysis and Classification Advanced demonstrate supervised
learning. Classification Analysis uses labelled training data to build a Neural
Network model to classify the activity type given the sensor data. ClassificationAdvanced
compares numerous classification techniques. The models are evaluated on a hold-out 
test set and shown to be able to classify activities with good accuracy. 

Required to run the examples:
MATLAB
Neural Network Toolbox (for the neural network portions)
Fuzzy Logic Toolbox (for the fuzzy c-mean algorithm)
Statistics Toolbox (for all other machine learning functionality)
Version R2014b is required for fitceoc function that automatically extends binary
classifiers such as SVM to multiclass problems.

Acknowledgements:
The dataset is courtesy:
Davide Anguita, Alessandro Ghio, Luca Oneto, Xavier Parra and Jorge L.
Reyes-Ortiz. Human Activity Recognition on Smartphones using a
Multiclass Hardware-Friendly Support Vector Machine. International
Workshop of Ambient Assisted Living (IWAAL 2012). Vitoria-Gasteiz,
Spain. Dec 2012

This example also uses the following MATLAB Central File Exchange entries:
http://www.mathworks.com/matlabcentral/fileexchange/27388-plot-and-compare-histograms--pretty-by-default
http://www.mathworks.com/matlabcentral/fileexchange/8577-scatplot

Copyright 2014-2015, The MathWorks, Inc.