Recommender systems have been based on context and content, and now the technological challenge of making personalized recommendations based on the user emotional state arises through physiological signals that are obtained from devices or sensors. This paper applies the deep learning approach using a deep convolutional neural network on a dataset of physiological signals (electrocardiogram and galvanic skin response), in this case, the AMIGOS dataset. The detection of emotions is done by correlating these physiological signals with the data of arousal and valence of this dataset, to classify the affective state of a person. In addition, an application for emotion recognition based on classic machine learning algorithms is proposed to extract the features of physiological signals in the domain of time, frequency, and non-linear. This application uses a convolutional neural network for the automatic feature extraction of the physiological signals, and through fully connected network layers, the emotion prediction is made. The experimental results on the AMIGOS dataset show that the method proposed in this paper achieves a better precision of the classification of the emotional states, in comparison with the originally obtained by the authors of this dataset. INDEX TERMS Emotion recognition, deep convolutional neural network, physiological signals, machine learning, AMIGOS dataset.

This paper presents an in-depth empirical analysis of a nine-week MOOC. This analysis provides novel results regarding participants' profiles and use of built-in and external social tools. The results served to detect seven participants' patterns and conclude that the forum was the social tool preferred to contribute to the MOOC.

No abstract

Objectives: Recognising human activity is very useful for an investigator about a patient's behaviour and can aid in prescribing activity in future recommendations. The use of body worn accelerometers has been demonstrated to be an accurate measure of human activity, however research looking at the use of multiple body worn accelerometers in a free living environment to recognise a wide range of activities is not evident. This study aimed to successfully recognise activity and subcategory activity types through the use of multiple body worn accelerometers in a free living environment. Method: Ten participants (Age = 23.1 ± 1.7 years, height =171.0 ± 4.7 cm, mass = 78.2 ± 12.5 Kg) wore nine body-worn accelerometers for a day of free living. Activity type was identified through the use of a wearable camera, and sub category activities were quantified through a combination of freeliving and controlled testing. A variety of machine learning techniques consisting of pre-processing algorithms, feature and classifier selections were tested, accuracy and computing time were reported. Results: A fine k-nearest neighbour classifier with mean and standard deviation features of unfiltered data reported a recognition accuracy of 97.6%. Controlled and free-living testing provided highly accurate recognition for sub-category activities (>95.0%). Decision tree classifiers and maximum features demonstrated to have the lowest computing time. Conclusions: Results show recognition of activity and subcategory activity types is possible in a free living environment through the use of multiple body worn accelerometers. This method can aid in prescribing recommendations for activity and sedentary periods for healthy living.

No abstract

Accurate estimations for the near future levels of blood glucose are crucial for Type 1 Diabetes Mellitus (T1DM) patients in order to be able to react on time and avoid hypo and hyper-glycemic episodes. Accurate predictions for blood glucose are the base for control algorithms in glucose regulating systems such as the artificial pancreas. Numerous research studies have already been conducted in order to provide predictions for blood glucose levels with particularities in the input signals and underlying models used. These models can be categorized into two major families: those based on tuning glucose physiological-metabolic models and those based on learning glucose evolution patterns based on machine learning techniques. This paper reviews the state of the art in blood glucose predictions for T1DM patients and proposes, implements, validates and compares a new hybrid model that decomposes a deep machine learning model in order to mimic the metabolic behavior of physiological blood glucose methods. The differential equations for carbohydrate and insulin absorption in physiological models are modeled using a Recurrent Neural Network (RNN) implemented using Long Short-Term Memory (LSTM) cells. The results show Root Mean Square Error (RMSE) values under 5 mg/dL for simulated patients and under 10 mg/dL for real patients.

Automatic detection of traffic lights, street crossings and urban roundabouts combining outlier detection and deep learning classification techniques based on GPS traces while driving.

Abstract:One of the most relevant difficulties faced by first-year undergraduate students is to settle into the educational environment of universities. This paper presents a case study that proposes a computer-assisted collaborative experience designed to help students in their transition from high school to university. This is done by facilitating their first contact with the campus and its services, the university community, methodologies and activities. The experience combines individual and collaborative activities, conducted in and out of the classroom, structured following the Jigsaw Collaborative Learning Flow Pattern. A specific environment including portable technologies with network and computer applications has been developed to support and facilitate the orchestration of a flow of learning activities into a single integrated learning setting. The result is a Computer Supported Collaborative Blended Learning scenario that has been evaluated with first-year university students of the degrees of Software and Audiovisual Engineering within the subject Introduction to Information and Communications Technologies. The findings reveal that the scenario improves significantly students' interest in their studies and their understanding about the campus and services provided. The environment is also an innovative approach to successfully support the heterogeneous activities conducted by both teachers and students during the scenario. This paper introduces the goals and context of the case study, describes how the technology was employed to conduct the learning scenario, the evaluation methods and the main results of the experience.