Mars: a personalised mobile activity recognition system

Mobile activity recognition focuses on inferring the current activities of a mobile user by leveraging the sensory data that is available on today’s smart phones. The state of the art in mobile activity recognition uses traditional classification learning techniques. Thus, the learning process typically involves: i) collection of labelled sensory data that is transferred and collated in a centralised repository; ii) model building where the classification model is trained and tested using the collected data; iii) a model deployment stage where the learnt model is deployed on-board a mobile device for identifying activities based on new sensory data. In this paper, we demonstrate the Mobile Activity Recognition System (MARS) where for the first time the model is built and continuously updated on-board the mobile device itself using data stream mining. The advantages of the on-board approach are that it allows model personalisation and increased privacy as the data is not sent to any external site. Furthermore, when the user or its activity profile changes MARS enables promptly adaptation. MARS has been implemented on the Android platform to demonstrate that it can achieve accurate mobile activity recognition. Moreover, we can show in practise that MARS quickly adapts to user profile changes while at the same time being scalable and efficient in terms of consumption of the device resources.

Análisis de la actividad física competitiva en jugadores de fútbol de un equipo de excelencia

OBJETIVO: Esta tesis doctoral pretende analizar el rendimiento en fútbol desde un punto de vista físico de los jugadores de un equipo de excelencia (EE). Se entiende como EE a un equipo que presenta un prolongado bagaje de éxitos nacionales e internacionales a lo largo de varios años. MÉTODO: Se utilizó el sistema de registro (AMISCO) de la actividad del jugador de fútbol en competición (distancia total recorrida, distancias a distintas velocidades, y número de sprints) analizando a lo largo de seis temporadas continuada la actividad de jugadores de élite (N=2082), siendo 1051 pertenecientes al EE y 1031 a jugadores de los equipos rivales (ER). RSULTADO: El perfil de actividad (distancia total, distancia en distintas velocidades y nº de sprints) de los jugadores del EE muestran valores inferiores al de los jugadores de los ER. De igual forma ocurre en cada una de las partes del partido, el puesto específico de juego de los jugadores en las distintas temporadas analizadas y en los distintos meses de la temporada. El EE registra una mayor actividad (distancia total recorrida en el partido, distancia recorrida en velocidades altas V>17 km/h y nº de Sprints) cuando juega de local a cuando lo hace como visitante. Asimismo EE muestra demandas superiores de exigencia física en la competición europea (Champions League) frente a la Liga en la distancia total recorrida en el partido y a velocidades bajas (V<17 km/h), mientras que apenas muestra diferencias en los indicadores de exigencia física cuando se enfrenta a los equipos rivales de distinto nivel de rendimiento, exceptuando en la distancia recorrida andando (V=01-11 km/h) con diferencias significativas superiores cuando el nivel del oponente es de Máxima dificultad frente a Elevada dificultad (p<.01) o Media dificultad (p<.01). En función del estilo de juego se registran diferencias en la distancia total recorrida en el partido, en velocidades bajas (V<17 km/h) y en el nº de Sprints (V>24 km/h). Por último, el perfil de actividad del EE no se ve influenciado por el efecto de estar sometido a una alta densidad competitiva (jugar un partido entre semana). Se constata como un jugador de un EE está supeditado no sólo a las exigencias del puesto específico sino también está sujeta a las características propias del jugador al desempeñar el rol o función táctica encomendada. Esta influencia es más evidente en jugadores que ocupan puestos específicos avanzados. CONCLUSION: La presente investigación aporta la caracterización de la actividad de los jugadores de un EE frente al perfil de actividad de sus rivales, mostrando un nivel de respuesta específica y diferente, más estable frente a determinadas variables contextuales y donde la exigencia física en el puesto específico se ve condicionada por las características individuales del jugador, todos estos aspectos pueden ayudar a la hora de planificar el entrenamiento en este tipo de equipos. ABSTRACT OBJETIVE: This doctoral thesis aims to analyze the performance in football from a physical point of view on a team of excellence. METHOD: The registration system (AMISCO) activity soccer player in competition (total distance traveled distances at different speeds, and number of sprints) was used to analyze over six seasons continued activity elite players (N = 2082), with 1051 belonging to Team Excellence (EE) and 1031 players from rival teams (ER). RESULT: The competitive physical response of the US players are analyzed and compared with that of ER. The activity profile (total distance, distance at different speeds and number of sprints) of EE players shown below players ER values. The same occurs in each of the parts of the game, the specific game since players analyzed in different seasons and in different months of the season. The EE recorded increased activity (total distance traveled in the match, distance at high speeds V> 17 km / h and sprints number) when playing a local when it does on the road. EE also shows higher demands physical demands in European competition (Champions League) against the League in the total distance traveled in the party and at low speeds (V <17 km / h), while just shows differences in indicators of demand physics when the opposing teams with different levels of performance faces, except in the distance walked (V = 01-11 km / h) with significant differences when higher level Maximum opponent is facing difficulty High difficulty (p < .01) or Medium difficulty (p <.01). Depending on the style of play differences are recognized in the total distance traveled in the party, at low speeds (V <17 km / h) and the number of Sprints (V> 24 km / h). Finally, the activity profile of EE is not influenced by the effect of being subjected to a high competitive density (play a game during the week). It is found as a player on a team of Excellence is subject not only to the requirements of the specific position but is also subject to the characteristics of the player to play the role or function entrusted tactics. This influence is most evident in advanced players who occupy specific positions. CONCLUSION: This research provides the characterization of the activity of EE players versus activity profile of its rivals, showing a level of response specific and different, more stable against certain contextual variables and where the physical demands on the specific position is conditioned by the individual characteristics of the player, all these aspects can help in planning training in this type of equipment.

Nutrition and physical activity programs for obesity treatment (PRONAF study). methodological approach of the project

Este trabajo aborda la metodología seguida para llevar a cabo el proyecto de investigación PRONAF (Clinical Trials Gov.: number NCT01116856.) Background: At present, scientific consensus exists on the multifactorial etiopatogenia of obesity. Both professionals and researchers agree that treatment must also have a multifactorial approach, including diet, physical activity, pharmacology and/or surgical treatment. These two last ones should be reserved for those cases of morbid obesities or in case of failure of the previous ones. The aim of the PRONAF study is to determine what type of exercise combined with caloric restriction is the most appropriate to be included in overweigth and obesity intervention programs, and the aim of this paper is to describe the design and the evaluation methods used to carry out the PRONAF study. Methods/design: One-hundred nineteen overweight (46 males) and 120 obese (61 males) subjects aged 18–50 years were randomly assigned to a strength training group, an endurance training group, a combined strength + endurance training group or a diet and physical activity recommendations group. The intervention period was 22 weeks (in all cases 3 times/wk of training for 22 weeks and 2 weeks for pre and post evaluation). All subjects followed a hypocaloric diet (25-30% less energy intake than the daily energy expenditure estimated by accelerometry). 29–34% of the total energy intake came from fat, 14–20% from protein, and 50–55% from carbohydrates. The mayor outcome variables assesed were, biochemical and inflamatory markers, body composition, energy balance, physical fitness, nutritional habits, genetic profile and quality of life. 180 (75.3%) subjects finished the study, with a dropout rate of 24.7%. Dropout reasons included: personal reasons 17 (28.8%), low adherence to exercise 3 (5.1%), low adherence to diet 6 (10.2%), job change 6 (10.2%), and lost interest 27 (45.8%). Discussion: Feasibility of the study has been proven, with a low dropout rate which corresponds to the estimated sample size. Transfer of knowledge is foreseen as a spin-off, in order that overweight and obese subjects can benefit from the results. The aim is to transfer it to sports centres. Effectiveness on individual health-related parameter in order to determine the most effective training programme will be analysed in forthcoming publications.

Effects of dietary restriction combined with different exercise programs or physical activity recommendations on blood lipids in overweight adults

Background and aim: Many exercise studies, although generally showing the beneficial effects of supervised aerobic, resistance or combined exercise on blood lipids, have sometimes reached equivocal conclusions. The aim of this study is to evaluate the impact of different programs that combined exercise and dietary restriction on blood lipids versus a clinical practice intervention for weight loss, in overweight adults. Methods: For this study 66 subjects participated in a supervised 22 weeks training program, composed of three sessions per week and they were randomized in three groups: strength training (S; n = 19), endurance training (E; n = 25), a combination of E and S (SE; n = 22). Eighteen subjects served as physical activity group (PA) that followed a clinical intervention consisted of physical activity recommendations. All groups followed the same dietary treatment, and blood samples were obtained for lipids measurements, at the beginning and end of the study. Results: Lipid profile improved in all groups. No significant differences for baseline and post-training values were observed between groups. In general, SE and PA decreased low-density lipoprotein cholesterol (LDL-C) values (p menor que 0.01). S decreased triglyceride levels (p menor que 0.01) and E, SE, and PA decreased total cholesterol levels (p menor que 0.05, p menor que 0.01 and p menor que 0.01, respectively). Conclusions: These results suggest that an intervention program of supervised exercise combined with diet restriction did not achieved further improvements in blood lipid profile than diet restriction and physical activity recommendations, in overweight adults. (Clinical Trials gov number: NCT01116856).