Vol.:(0123456789) 1 3
Journal of Ambient Intelligence and Humanized Computing
https://doi.org/10.1007/s12652-020-02351-x
ORIGINAL RESEARCH
Human activity recognition based on smartphone using fast feature
dimensionality reduction technique
B. A. Mohammed Hashim
1
· R. Amutha
2
Received: 7 January 2020 / Accepted: 17 July 2020
© Springer-Verlag GmbH Germany, part of Springer Nature 2020
Abstract
Human activity recognition aims to identify the activities carried out by a person. Recognition is possible by using informa-
tion that is retrieved from numerous physiological signals by attaching sensors to the subject’s body. Lately, sensors like
accelerometer and gyroscope are built-in inside the Smartphone itself, which makes activity recognition very simple. To
make the activity recognition system work properly in smartphone which has power constraint, it is very essential to use an
optimization technique which can reduce the number of features used in the dataset with less time consumption. In this paper,
we have proposed a dimensionality reduction technique called fast feature dimensionality reduction technique (FFDRT).
A dataset (UCI HAR repository) available in the public domain is used in this work. Results from this study shows that
the fast feature dimensionality reduction technique applied for the dataset has reduced the number of features from 561 to
66, while maintaining the activity recognition accuracy at 98.72% using random forest classifer and time consumption in
dimensionality reduction stage using FFDRT is much below the state of the art techniques.
Keywords Activity recognition · Smartphone · Accelerometer · FFDRT · Random forest
1 Introduction
Since 1979, the year in which the frst ever commercial
handheld mobile phones were introduced, there has been
no stopping in the growth of it. The mobile phones were
used by 80% of the world population by the year 2011. In a
very short period of time, mobile phones have become part
of our life. The new generation smartphones are equipped
with many features with the help of many sensors. With
the use of these features in the smartphones, it is very easy
to keep track of our daily activities. These kinds of assis-
tive technologies can be useful for remote health care, for
the disabled, the elderly and to those with important needs.
Human activity recognition is a very active research feld
in which techniques for understanding human activities by
interpreting attributes taken from motion, physiological sig-
nals, location and environmental information etc.,
Human activity recognition identifes the actions done
by a person given a set of observations. The recognition
can be done by using the information taken from sensors.
The main objective of activity recognition system is to iden-
tify the actions performed by humans, from the collected
data through sensors. The latest smartphones have motion,
acceleration or inertial sensors, and by using the information
taken from the sensors, recognition of activities is possible.
Since it is possible to collect a large amount of data through
the features available in smartphones but the challenge lies
in processing the data and implementing it in real time appli-
cations. So, there is an important need for new data mining
schemes to end these kinds of challenges. Since we have
many features derived from sensors like accelerometer and
gyroscope, it is essential to use feature selection algorithms
to simplify the computing process by reducing the number of
features. mRMR feature selection algorithm has been used
in a study to reduce the dimensionality of features (Doewes
2017). Accordingly, in this paper fast feature dimensionality
reduction technique has been employed for dimensionality
* B. A. Mohammed Hashim
hashimba.ece@cahcet.edu.in
R. Amutha
amuthar@ssn.edu.in
1
Department of Electronics and Communication Engineering,
C. Abdul Hakeem College of Engineering and Technology,
Hakeem Nagar, Melvisharam, Ranipet, Tamil Nadu, India
2
Department of Electronics and Communication Engineering,
SSN College of Engineering, Kalavakkam, Chennai,
Tamil Nadu, India