Support Vector Machine for Behavior-Based Driver Identification System

Abstract

We present an intelligent driver identification system to handle vehicle theft based on modeling dynamic human behaviors. We propose to recognize illegitimate drivers through their driving behaviors. Since human driving behaviors belong to a dynamic biometrical feature which is complex and difficult to imitate compared with static features such as passwords and fingerprints, we find that this novel idea of utilizing human dynamic features for enhanced security application is more effective. In this paper, we first describe our experimental platform for collecting and modeling human driving behaviors. Then we compare fast Fourier transform (FFT), principal component analysis (PCA), and independent component analysis (ICA) for data preprocessing. Using machine learning method of support vector machine (SVM), we derive the individual driving behavior model and we then demonstrate the procedure for recognizing different drivers by analyzing the corresponding models. The experimental results of learning algorithms and evaluation are described.

Combining Unsupervised Anomaly Detection and Neural Networks for Driver Identification

“…Using real-world data, it is shown that only seven-second-long driving behavior signals and geolocation signals are used to identify drivers. Furthermore, the proposed algorithm achieves relatively high accuracy and also outperforms the previously proposed algorithms [6,7,9,10]. Lastly, for implementation purposes, real-world results associated with each operation of the proposed algorithm are shown step-by-step.…”

“…More recently proposed algorithms use Support Vector Machine (SVM) and are more focused in using driving behavior signals as inputs [10,16]. In [16], signals from inertial sensors, namely, accelerometer and gyroscope, are processed in order to detect accelerating, braking, and wheel-turning events.…”

“…Therefore, for benchmarking, we select three previously proposed algorithms, namely, Benchmark 1 which is our previously proposed algorithm in [6], Benchmark 2 which is Gaussian Mixture Model (GMM) [7,9], and Benchmark 3 which is based on Support Vector Machine (SVM) [10]. Benchmark 1 is chosen because it is designed specifically to use similar input signals as the algorithm proposed in this paper [6].…”

“…Support Vector Machine (SVM) is chosen as Benchmark 3 because it often constitutes the core of recently proposed driver identification algorithms as well as in other classification problems [10,16]. In SVM-based algorithms, an optimal hyperplane is found to separate drivers using support vector methods.…”

“…In SVM-based algorithms, an optimal hyperplane is found to separate drivers using support vector methods. For driver identification, the algorithms proposed in [10] are based on SVM and use driving behavior signals including acceleration and braking as their primary input signals. We note that all three benchmark algorithms are assessed using the same set of data as the proposed algorithm.…”