| 1. |
- Mitra, Karan, et al.
(author)
-
ALPINE : A Bayesian System For Cloud Performance Diagnosis And Prediction
- 2017
-
In: 2017 IEEE International Conference on Services Computing (SCC). - Piscataway, NJ : IEEE. - 9781538620052 ; , s. 281-288
-
Conference paper (peer-reviewed)abstract
- Cloud performance diagnosis and prediction is a challenging problem due to the stochastic nature of the cloud systems. Cloud performance is affected by a large set of factors such as virtual machine types, regions, workloads, wide area network delay and bandwidth. Therefore, necessitating the determination of complex relationships between these factors. The current research in this area does not address the challenge of modeling the uncertain and complex relationships between these factors. Further, the challenge of cloud performance prediction under uncertainty has not garnered sufficient attention. This paper proposes, develops and validates ALPINE, a Bayesian system for cloud performance diagnosis and prediction. ALPINE incorporates Bayesian networks to model uncertain and complex relationships between several factors mentioned above. It handles missing, scarce and sparse data to diagnose and predict stochastic cloud performance efficiently. We validate our proposed system using extensive real data and show that it predicts cloud performance with high accuracy of 91.93%.
|
|
| 2. |
- Nugent, Christopher, et al.
(author)
-
Improving the Quality of User Generated Data Sets for Activity Recognition
- 2016
-
In: Ubiquitous Computing and Ambient Intelligence, UCAMI 2016, PT II. - Amsterdam : Springer Publishing Company. - 9783319487991 - 9783319487984 ; , s. 104-110
-
Conference paper (peer-reviewed)abstract
- It is fully appreciated that progress in the development of data driven approaches to activity recognition are being hampered due to the lack of large scale, high quality, annotated data sets. In an effort to address this the Open Data Initiative (ODI) was conceived as a potential solution for the creation of shared resources for the collection and sharing of open data sets. As part of this process, an analysis was undertaken of datasets collected using a smart environment simulation tool. A noticeable difference was found in the first 1-2 cycles of users generating data. Further analysis demonstrated the effects that this had on the development of activity recognition models with a decrease of performance for both support vector machine and decision tree based classifiers. The outcome of the study has led to the production of a strategy to ensure an initial training phase is considered prior to full scale collection of the data.
|
|
| 3. |
- Balouji, Ebrahim, 1985, et al.
(author)
-
A LSTM-based Deep Learning Method with Application to Voltage Dip Classification
- 2018
-
In: 2018 18TH INTERNATIONAL CONFERENCE ON HARMONICS AND QUALITY OF POWER (ICHQP). - Piscataway, NJ : Institute of Electrical and Electronics Engineers (IEEE). - 2164-0610. - 9781538605172 - 9781538605172
-
Conference paper (peer-reviewed)abstract
- In this paper, a deep learning (DL)-based method for automatic feature extraction and classification of voltage dips is proposed. The method consists of a dedicated architecture of Long Short-Term Memory (LSTM), which is a special type of Recurrent Neural Networks (RNNs). A total of 5982 three-phase one-cycle voltage dip RMS sequences, measured from several countries, has been used in our experiments. Our results have shown that the proposedmethod is able to classify the voltage dips from learned features in LSTM, with 93.40% classification accuracy on the test data set. The developed architecture is shown to be novel for feature learning and classification of voltage dips. Different from the conventional machine learning methods, the proposed method is able to learn dip features without requiring transition-event segmentation, selecting thresholds, and using expert rules or human expert knowledge, when a large amount of measurement data is available. This opens a new possibility of exploiting deep learning technology for power quality data analytics and classification.
|
|
| 4. |
- Heintz, Fredrik, et al.
(author)
-
Computing at School in Sweden - Experiences from Introducing Computer Science within Existing Subjects
- 2015
-
In: Informatics in Schools. Curricula, Competences, and Competitions /Lecture Notes in Computer Science and General Issues. - Cham : Springer International Publishing. - 1611-3349 .- 0302-9743. - 9783319253954 - 9783319253961 ; 9378, s. 118-130
-
Conference paper (peer-reviewed)abstract
- Computing is no longer considered a subject area only relevant for a narrow group of professionals, but rather as a vital part of general education that should be available to all children and youth. Since making changes to national curricula takes time, people are trying to find other ways of introducing children and youth to computing. In Sweden, several current initiatives by researchers and teachers aim at finding ways of working with computing within the current curriculum. In this paper we present case studies based on a selection of these initiatives from four major regions in Sweden and based on these case studies we present our ideas for how to move forward on introducing computational thinking on a larger scale in Swedish education.
|
|
| 5. |
- Heintz, Fredrik, 1975-, et al.
(author)
-
Introducing Programming and Digital Competence in Swedish K-9 Education
- 2017
-
In: Informatics in Schools. - Cham : Springer. - 1611-3349 .- 0302-9743. - 9783319714820 - 9783319714837 ; , s. 117-128
-
Conference paper (peer-reviewed)abstract
- The role of computer science and IT in Swedish schools has varied throughout the years. In fall 2014, the Swedish government gave the National Agency for Education (Skolverket) the task of preparing a proposal for K–9 education on how to better address the competences required in a digitalized society. In June 2016, Skolverket handed over a proposal introducing digital competence and programming as interdisciplinary traits, also providing explicit formulations in subjects such as mathematics (programming, algorithms and problem-solving), technology (controlling physical artifacts) and social sciences (fostering aware and critical citizens in a digital society). In March 2017, the government approved the new curriculum, which needs to be implemented by fall 2018 at the latest. We present the new K–9 curriculum and put it in a historical context. We also describe and analyze the process of developing the revised curriculum, and discuss some initiatives for how to implement the changes.
|
|
| 6. |
- Nordlander, Johan, et al.
(author)
-
Unambiguous Semantics In Automotive Timing Modeling
- 2010
-
In: Proceedings of the 1st Workshop on Critical Automotive applications: Robustness & Safety (CARS '10) in Conjunction With The 8th European Dependable Computing Conference (EDCC-8), Valencia, Spain, 2010.. - New York, New York, USA : ACM Press. - 9781605589152 ; , s. 39-42
-
Conference paper (peer-reviewed)abstract
- This paper shows how the ITEA2 research project TIMMO has chosen to model timing constraints for delays and for synchronization in an unambiguous way. The shown timing model is valid even in systems where jitter and over- and under-sampling appears. These timing constraints can be used to model timing by augmenting both the AUTOSAR and EAST-ADL2 languages. The unambiguous semantics enables the use of timing models for building safety cases and for precise communication between different parties involved in safety-critical automotive applications. © 2010 ACM.
|
|
| 7. |
- Greef, T.J., et al.
(author)
-
Connected media and presence
- 2013
-
In: SAM 2013. - New York, NY, USA : ACM Press. - 9781450323949 ; , s. 43-48
-
Conference paper (peer-reviewed)abstract
- Effective design of shared mediated spaces, information and connectedness requires theory and practice from a range of disciplines such as found in European projects like Together Anywhere, Together Anytime (TA2) and the EIT ICT Labs Mediating Presence activity. Building on this work we continue to investigate the changes in the European digital media industry such as changed traditional distribution of media content and the progressive integration of (social) communication means in information distribution and shared mediated spaces. Our past research has given valuable insights in how to design and evaluate systems and services that provide a high quality of experience, in how trust is established in mediated environments, and how the formation of tacit communication between participants in new distributed and connected media is negotiated. In the new Seventh Framework Program project COnnected Media and Presence from European Institute of Technology (COMPEIT) we aim to enhance the quality of experience in face-to-face and broadcast communication further in three domains: 1) Spatial connectedness, 2) Social connectedness and 3) Information connectedness, by developing three key services: Shared Experience with Tangible Interaction (SETI); Broadcast Presence Studio (BPS) and Mixed-Reality Interaction (MRI). The quality of experience of these services will be enhanced in terms of for example: spatial connectedness, by providing shared spaces and supporting spatial features such as eye contact; social connectedness, by using natural means for interaction suiting different settings and activities; and information connectedness, by providing better means to share, manipulate and use information suiting different task or activities. In this paper we will discuss the background of this work and give an overview of our planned future work in COMPEIT.
|
|
| 8. |
- Alawadi, Sadi, 1983-, et al.
(author)
-
A Federated Interactive Learning IoT-Based Health Monitoring Platform
- 2021
-
In: New Trends in Database and Information Systems. - Cham : Springer. ; , s. 235-246, s. 235-246
-
Conference paper (peer-reviewed)abstract
- Remote health monitoring is a trend for better health management which necessitates the need for secure monitoring and privacy-preservation of patient data. Moreover, accurate and continuous monitoring of personal health status may require expert validation in an active learning strategy. As a result, this paper proposes a Federated Interactive Learning IoT-based Health Monitoring Platform (FIL-IoT-HMP) which incorporates multi-expert feedback as ‘Human-in-the-loop’ in an active learning strategy in order to improve the clients’ Machine Learning (ML) models. The authors have proposed an architecture and conducted an experiment as a proof of concept. Federated learning approach has been preferred in this context given that it strengthens privacy by allowing the global model to be trained while sensitive data is retained at the local edge nodes. Also, each model’s accuracy is improved while privacy and security of data has been upheld.
|
|
| 9. |
- Alonso, Pedro, 1986-, et al.
(author)
-
HyperEmbed: Tradeoffs Between Resources and Performance in NLP Tasks with Hyperdimensional Computing Enabled Embedding of n-gram Statistics
- 2021
-
In: 2021 International Joint Conference on Neural Networks (IJCNN) Proceedings. - : IEEE.
-
Conference paper (peer-reviewed)abstract
- Recent advances in Deep Learning have led to a significant performance increase on several NLP tasks, however, the models become more and more computationally demanding. Therefore, this paper tackles the domain of computationally efficient algorithms for NLP tasks. In particular, it investigates distributed representations of n -gram statistics of texts. The representations are formed using hyperdimensional computing enabled embedding. These representations then serve as features, which are used as input to standard classifiers. We investigate the applicability of the embedding on one large and three small standard datasets for classification tasks using nine classifiers. The embedding achieved on par F1 scores while decreasing the time and memory requirements by several times compared to the conventional n -gram statistics, e.g., for one of the classifiers on a small dataset, the memory reduction was 6.18 times; while train and test speed-ups were 4.62 and 3.84 times, respectively. For many classifiers on the large dataset, memory reduction was ca. 100 times and train and test speed-ups were over 100 times. Importantly, the usage of distributed representations formed via hyperdimensional computing allows dissecting strict dependency between the dimensionality of the representation and n-gram size, thus, opening a room for tradeoffs.
|
|
| 10. |
|
|
