Eylem Seç
An Empirical Evaluation of Machine Learning Classification Algorithms on User Movement Data
Başlık:
An Empirical Evaluation of Machine Learning Classification Algorithms on User Movement Data
Yazar:
Kelley, Christopher, author.
ISBN:
9780355983937
Yazar Ek Girişi:
Fiziksel Tanımlama:
1 electronic resource (55 pages)
Genel Not:
Source: Masters Abstracts International, Volume: 57-06M(E).
Advisors: Kaushik Roy; Dorothy Yuan Committee members: Albert Esterline; Kaushik Roy; Dorothy Yuan.
Özet:
In the past few years we have seen a big growth in the research of biometrics. We see biometrics implemented everywhere especially in security. Almost every mobile device today uses biometrics as an authentication method, we see this in the form of fingerprint sensors, Iris scanners, and even facial scanners. As Moore's law continues to hold true, our mobile devices are becoming even more powerful. These powerful mobile devices coupled with machine learning gives us the opportunity to change the way we use them in our everyday lives. People are constantly using their mobile devices to track everything from nutrition and sleep, to fitness and their own movements in general. This movement data can be collected and used to add new security functionality to user's mobile devices. Measuring a user's movement using mobile devices allows for the use of behavioral biometrics. This could induce a shift in our methods of securing mobile devices away from physical attributes like fingerprints or our face towards behavioral attributes like the way we walk or perform some personal activity.
In this thesis, an empirical evaluation of different classification techniques is conducted on user movement data. The datasets used during experimentation are composed of accelerometer data collected from various mobile devices, including smartphones, smart watches, and other accelerometer sensors. The user movement data was processed and fed into five traditional machine-learning algorithms. Their classification performances were then compared with a deep learning technique's, the Long Short Term Memory-Recurrent Neural Network (LSTM-RNN). LSTM-RNN achieved its highest accuracy at 89% opposed to 97% from a traditional machine-learning algorithm, specifically the k-Nearest Neighbor (k-NN) on wrist-worn accelerometer data.
Notlar:
School code: 1544
Konu Başlığı:
Tüzel Kişi Ek Girişi:
Mevcut:*
Yer Numarası | Demirbaş Numarası | Shelf Location | Lokasyon / Statüsü / İade Tarihi |
---|---|---|---|
XX(689709.1) | 689709-1001 | Proquest E-Tez Koleksiyonu | Arıyor... |
On Order
Liste seç
Bunu varsayılan liste yap.
Öğeler başarıyla eklendi
Öğeler eklenirken hata oldu. Lütfen tekrar deneyiniz.
:
Select An Item
Data usage warning: You will receive one text message for each title you selected.
Standard text messaging rates apply.