e-journal

An Unsupervised Approach for Automatic Activity Recognition Based on Hidden Markov Model Regression

---

Using supervised machine learning approaches to recognize human activities from on-body wearable accelerometers generally requires a large amount of labeled data.When ground truth information is not available, too expensive, time consuming or difficult to collect, one has to rely on unsupervised approaches. This paper presents a new unsupervised approach for human activity recognition from raw acceleration data measured using inertial wearable sensors. The proposed method is based upon
joint segmentation of multidimensional time series using a HiddenMarkov Model (HMM) in a multiple regression context. The model is learned in an unsupervised framework using the Expectation-Maximization (EM) algorithm where no activity labels are needed. The proposed method takes
into account the sequential appearance of the data. It is therefore adapted for the temporal acceleration data to accurately detect the activities. It allows both segmentation and classification of the human activities. Experimental results are provided to demonstrate the efficiency of the proposed
approach with respect to standard supervised and unsupervised classification approaches.
Note to Practitioners—This paper was motivated by the problem of automatic recognition of physical human activities using on-body wearable sensors in a health-monitoring context. Threewearable sensors (accelerometers) are placed at the chest, the right thigh and the left ankle of the subject.
The studied activities concern the static ones, the dynamic ones as well as transitions between those activities. Since the goal is to recognize human activities from only the raw acceleration data, the acquired acceleration signals are seen as multidimensional time series with regime changes due to the changes of activities over time. The activity recognition problem is formulated as a problem of multidimensional time series segmentation. Segmenting the time series according to different unknown regimes over time is equivalent to classifying the acceleration data into one set of activities;
each activity being associated with a regime. The proposed approach does not require any annotation of the raw accelerations by experts to learn the model parameters. However, it assumes that the number of activities is known. This assumption can be a constraint in a context of exploratory data mining where the aim is to automatically cluster a large amount of data into different group of activities. To tackle this limitation, a selection criterion can be used to determine the number of the groups of activities.
Index Terms—Activity recognition, hiddenMarkov model (HMM), multivariate regression, unsupervised learning, wearable computing.

---

Ketersediaan

Tidak ada salinan data

Informasi Detail

Judul Seri

IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING

No. Panggil

-

Penerbit

New York : IEEE., 2013

Deskripsi Fisik

IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING, VOL. 10, NO. 3, JULY 2013

Bahasa

English

ISBN/ISSN

-

Klasifikasi

-

Tipe Isi

-

Tipe Media

-

Tipe Pembawa

-

Edisi

VOL. 10, NO. 3

Subjek

TEKNIK

Info Detail Spesifik

-

Pernyataan Tanggungjawab

yuli/agus

Versi lain/terkait

Lampiran Berkas

Komentar

---

Anda harus masuk sebelum memberikan komentar