Loading...
10 results
Search Results
Now showing 1 - 10 of 10
- Model combination in neural-based forecastingPublication . Freitas, Paulo S. A.; Rodrigues, António J. L.This paper discusses different ways of combining neural predictive models or neural-based forecasts. The proposed approaches consider Gaussian radial basis function networks, which can be efficiently identified and estimated through recursive/adaptive methods. The usual framework for linearly combining estimates from different models is extended, to cope with the case where the forecasting errors from those models are correlated. A prefiltering methodology is pro posed, addressing the problems raised by heavily nonstationary time series. Moreover, the paper discusses two approaches for decision-making from forecasting models: either inferring decisions from combined predictive estimates, or combining prescriptive solutions derived from different forecasting models.
- The combination of neural estimates in prediction and decision problemsPublication . Freitas, Paulo Sérgio AbreuIn this dissertation, different ways of combining neural predictive models or neural-based forecasts are discussed. The proposed approaches consider mostly Gaussian radial basis function networks, which can be efficiently identified and estimated through recursive/adaptive methods. Two different ways of combining are explored to get a final estimate – model mixing and model synthesis –, with the aim of obtaining improvements both in terms of efficiency and effectiveness. In the context of model mixing, the usual framework for linearly combining estimates from different models is extended, to deal with the case where the forecast errors from those models are correlated. In the context of model synthesis, and to address the problems raised by heavily nonstationary time series, we propose hybrid dynamic models for more advanced time series forecasting, composed of a dynamic trend regressive model (or, even, a dynamic harmonic regressive model), and a Gaussian radial basis function network. Additionally, using the model mixing procedure, two approaches for decision-making from forecasting models are discussed and compared: either inferring decisions from combined predictive estimates, or combining prescriptive solutions derived from different forecasting models. Finally, the application of some of the models and methods proposed previously is illustrated with two case studies, based on time series from finance and from tourism.
- Yarn: a product for unraveling storiesPublication . Buenafe, Susan; Guzman, Luis; Kannan, Namrata; Mendoza, Kristine; Nunes, Nuno Jardim; Nisi, Valentina; Campos, Pedro; Gonçalves, Frederica; Campos, Miguel; Freitas, PauloWriting is one of the oldest human activities, dating back as far as 3200 BCE. This paper provides an industrial case study about understanding the creative writing process using interviews and directed storytelling on aspiring and established writers and educators, performed during a one year capstone project, where teams of HCI students pair up with industrial designers and developers in order to solve a real world design problem. After 26 interviews and 55 hours of analysis, four concepts were used as dimensions to analyse creative writing applications: serendipity, haven, evolution and shuffle. Based on these ideas, we developed a series of prototypes by gradually increasing the fidelity of each successive prototype and making changes elicited from user feedback. The culmination of our process is Yarn, a new writing application. Yarn helps writers “unravel their story.” With Yarn, a writer can (i) Play with structure; (ii) Easily move chunks of writing; (ii) Create alternatives of sections, and (iv) Write in a beautiful distraction-free way.
- Erratum to: Foot-turistic multimedia: designing interactive multimedia installations for shoe shopsPublication . Campos, Pedro; Campos, Miguel; Freitas, Paulo; Jorge, Joaquim
- SENSE-SEAT: reimagining ergonomics for a creativity support workstationPublication . Campos, Pedro; Pestana, João; Campos, Miguel; Freitas, Paulo; Ehrenberg, Nils; Hidzik, WojciechWe present an approach for improving creativity at the work place based on reimagining office furniture and bringing sub tle technological elements to persuade office workers towards more healthy, creative workstyles. SENSE-SEAT aims to shed light on how we can better design interactive furniture for the workplace.
- Foot-turistic multimedia: designing interactive multimedia installations for shoe shopsPublication . Campos, Pedro; Campos, Miguel; Freitas, Paulo; Jorge, JoaquimThe amount of money spent in a store is positively correlated with the amount of time spent inside. We argue this is an opportunity for multimedia installations that can entertain shoppers and promote interaction with the shop’s products. This was the main principle behind our design idea for two interactive installations specifically conceived for shoe shops. We present two applications of interactive multimedia to shoe shopping: an interactive semantic mirror and an interactive window logo. We also describe the results of ethnographic studies, before and after the design process. Our contribution is two-fold: (i) we develop and apply a new multimedia architecture that combines RFID in-store technology with high-end motion detection algorithms, and (ii) we describe one of the first few studies about multimedia installations for improving the shoe shopping experience, in what we call “foot-turistic” interactions.
- Second look: combining interactive surfaces with wearable computing to support creative writingPublication . Campos, Pedro; Gonçalves, Frederica; Martins, Michael; Campos, Miguel; Freitas, PauloWe present "Second Look", a platform of interactive surfaces and wearable computing for helping people, in particular creative writers, to overcome writer's block. The novelty of our systems stems from the addition of wearable devices (Google Glass) and crowdsourcing to improve creative writing on tablets and phones. A primary challenge in developing and evaluating creativity support tools is that we are not able to detect when a person is being creative. Our approach improves current ones by exploring the "in-the-moment" creativity and supporting it with adaptive ubiquitous technologies that try to keep people in a creative experience peak for a longer period of time.
- EmC-ICDSST 2019: 5th International Conference on Decision Support System Technology - ICDSST 2019 & EURO Mini Conference 2019 on "Decision Support Systems: Main Developments & Future Trends"Publication . Freitas, P. S. Abreu; Dargam, F.; Ribeiro, R.; Moreno Jimenez, J. M.; Papathanasiou, J.
- Second look: combining wearable computing and crowdsourcing to support creative writingPublication . Campos, Pedro; Gonçalves, Frederica; Martins, Michael; Campos, Miguel; Freitas, PauloWe present “Second Look”, a platform for helping people, in particular creative writers, to overcome writer’s block. This ubiquitous platform combines augmented reality (Google Glass and AR markers), ubiquitous computing (mobile phones), and crowdsourcing in order to improve the creativity, focus and performance of creative writers. A primary challenge in developing and evaluating creativity support tools is that we are not able to detect when a person is being creative. Our approach improves current ones by exploring the “in-the moment” creativity and supporting it with adaptive ubiquitous technologies that try to keep people in a creative experience peak for a longer period of time.
- Experiments in machine learning of alpha-decay half-livesPublication . Freitas, Paulo S. A.; Clark, John W.Artificial neural networks are trained by a standard backpropagation learning algorithm with regularization to model and predict the systematics of alpha decay of heavy and superheavy nuclei. This approach to regression is implemented in two alternative modes: (i) construction of a statistical global model based solely on available experimental data for alpha-decay half-lives and Q-values, and (ii) modeling of the residuals between the predictions of state-of-the-art phenomenological model (specifically, the effective liquid-drop model (ELDM)) and experiment. Analysis of the results provides insights on the strengths and limitations of this application of machine learning (ML) to exploration of the nuclear landscape in regions beyond the valley of stability.