Suchergebnisse

International audience ; The recent development of Internet of Things (IoT) has democratized activity monitoring. Even if the data collected can be useful for healthcare, sharing this sensitive information exposes users to privacy threats and re-identification. This paper presents two approaches to anonymize the motion sensor data. The first is an extension of an earlier work based on filtering in the time-frequency plane and convolutional neural network; and the second is based on handcrafted features extracted from the zeros distribution of the time-frequency representation. The two approaches are evaluated on a public dataset to assess the accuracy of activity recognition and user re-identification. With the first approach we obtained an accuracy rate in activity recognition of 73% while limiting the identity recognition to an accuracy rate of 30% which corresponds to an activity identity ratio of 2.4. With the second approach we succeeded in improving the activity and identity ratio to 2.67 by attaining an accuracy rate in activity recognition of 80% while maintaining the re-identification rate at 30%.

International audience ; The recent development of Internet of Things (IoT) has democratized activity monitoring. Even if the data collected can be useful for healthcare, sharing this sensitive information exposes users to privacy threats and re-identification. This paper presents two approaches to anonymize the motion sensor data. The first is an extension of an earlier work based on filtering in the time-frequency plane and convolutional neural network; and the second is based on handcrafted features extracted from the zeros distribution of the time-frequency representation. The two approaches are evaluated on a public dataset to assess the accuracy of activity recognition and user re-identification. With the first approach we obtained an accuracy rate in activity recognition of 73% while limiting the identity recognition to an accuracy rate of 30% which corresponds to an activity identity ratio of 2.4. With the second approach we succeeded in improving the activity and identity ratio to 2.67 by attaining an accuracy rate in activity recognition of 80% while maintaining the re-identification rate at 30%.

International audience ; The recent development of Internet of Things (IoT) has democratized activity monitoring. Even if the data collected can be useful for healthcare, sharing this sensitive information exposes users to privacy threats and re-identification. This paper presents two approaches to anonymize the motion sensor data. The first is an extension of an earlier work based on filtering in the time-frequency plane and convolutional neural network; and the second is based on handcrafted features extracted from the zeros distribution of the time-frequency representation. The two approaches are evaluated on a public dataset to assess the accuracy of activity recognition and user re-identification. With the first approach we obtained an accuracy rate in activity recognition of 73% while limiting the identity recognition to an accuracy rate of 30% which corresponds to an activity identity ratio of 2.4. With the second approach we succeeded in improving the activity and identity ratio to 2.67 by attaining an accuracy rate in activity recognition of 80% while maintaining the re-identification rate at 30%.

International audience ; The recent development of Internet of Things (IoT) has democratized activity monitoring. Even if the data collected can be useful for healthcare, sharing this sensitive information exposes users to privacy threats and re-identification. This paper presents two approaches to anonymize the motion sensor data. The first is an extension of an earlier work based on filtering in the time-frequency plane and convolutional neural network; and the second is based on handcrafted features extracted from the zeros distribution of the time-frequency representation. The two approaches are evaluated on a public dataset to assess the accuracy of activity recognition and user re-identification. With the first approach we obtained an accuracy rate in activity recognition of 73% while limiting the identity recognition to an accuracy rate of 30% which corresponds to an activity identity ratio of 2.4. With the second approach we succeeded in improving the activity and identity ratio to 2.67 by attaining an accuracy rate in activity recognition of 80% while maintaining the re-identification rate at 30%.

International audience ; The recent development of Internet of Things (IoT) has democratized activity monitoring. Even if the data collected can be useful for healthcare, sharing this sensitive information exposes users to privacy threats and re-identification. This paper presents two approaches to anonymize the motion sensor data. The first is an extension of an earlier work based on filtering in the time-frequency plane and convolutional neural network; and the second is based on handcrafted features extracted from the zeros distribution of the time-frequency representation. The two approaches are evaluated on a public dataset to assess the accuracy of activity recognition and user re-identification. With the first approach we obtained an accuracy rate in activity recognition of 73% while limiting the identity recognition to an accuracy rate of 30% which corresponds to an activity identity ratio of 2.4. With the second approach we succeeded in improving the activity and identity ratio to 2.67 by attaining an accuracy rate in activity recognition of 80% while maintaining the re-identification rate at 30%.

Los cuatro trabajos de investigación redactados en esta tesis doctoral intentan comprender mejor el comportamiento del mercado petrolero. En el primer capítulo de la tesis, el objetivo y enfoque ha sido analizar primero las propiedades estadísticas de los precios reales del petróleo utilizando las raíces unitarias y los métodos de integración fraccional y luego entender el comportamiento de los precios del petróleo antes y después de conflictos militares y eventos geopolíticos producidos desde la Segunda Guerra Mundial. En el segundo y tercer trabajo, contribuyo a la literatura sobre el comportamiento del precio del petróleo crudo y examino cómo este afecta a las fusiones y adquisiciones en la industria petrolera en los Estados Unidos. Analizamos la relación de estas dos series estudiando su dinámica en el tiempo utilizando datos de enero de 1980 a junio de 2012, respectivamente. En el segundo trabajo, empleamos metodologías basadas en la integración fraccional y la cointegración. El tercer trabajo, emplea una metodología basada en la transformada wavelet. El cuarto capítulo analiza la relación entre la producción total de petróleo crudo de los Estados Unidos (incluida la producción de petróleo de esquisto) y los precios del crudo WTI mediante el estudio en el dominio de tiempo-frecuencia utilizando una metodología basada en la transformada wavelet.