Driving Behavior Classification Using Smartphone Sensor Data
8th International Conference on Computer Science and Engineering, UBMK 2023, Burdur, Türkiye, 13 - 15 Eylül 2023, ss.370-375, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Doi Numarası: 10.1109/ubmk59864.2023.10286698
- Basıldığı Şehir: Burdur
- Basıldığı Ülke: Türkiye
- Sayfa Sayıları: ss.370-375
- Anahtar Kelimeler: behavior, classification, driving, safety, sensor
- Boğaziçi Üniversitesi Adresli: Evet
Özet
Given the substantial number of accidents because of human error, the development of automatic driver behavior monitoring systems has become a pressing need. By providing real-time monitoring and analysis of driving behavior, these systems have the potential to decrease the incidence of accidents and improve overall road safety. This study presents a novel analysis framework for classifying driving behavior based on data gathered from passengers' smartphones. The data were collected using our mobile application installed on the smartphones and processed using machine learning algorithms. The study utilized several machine learning classification t echniques, w ith a focus on developing a Long Short Term Memory (LSTM) algorithm for improved accuracy. A Federated Learning algorithm was also developed to collect the data not into one area and train a global model to apply for labeling the driving behavior. The results show the efficacy of the proposed approach in accurately classifying driving behavior based on data obtained from smartphones.