Pocket Data Mining

Gaber, Mohamed Medhat; Stahl, Frederic; Gomes, João Bártolo

doi:10.1007/978-3-319-02711-1

Cited by 25 publications

(6 citation statements)

References 70 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Formerly core functionalities such as phone calls and text messages, seem nowadays just like an additional feature rather than core functions. Such mobile data mining systems started off with basic Mobile Interfaces executing data analytics tasks server side but then quickly moved into on-board execution and hybrids [130].…”

Section: ) Data Stream Mining On the Go In Edge Environmentsmentioning

confidence: 99%

“…As smartphones became available capable of recording and processing data in real-time through their sensor and processing capabilities, software systems that allow execution of algorithms on smartphones have emerged, like the Open Mobile Miner (OMM) [134]. Shortly after the Pocket Data Mining System (PDM) framework appeared as a first proof of concept that exploration and collaborative data mining is possible in streaming environments [130], [135]. Since PDM various smartphone based data stream mining technologies emerged, such as Mobile Sensor Data Engine (MOSDEN) [136].…”

Section: ) Data Stream Mining On the Go In Edge Environmentsmentioning

confidence: 99%

See 1 more Smart Citation

Mapping the Big Data Landscape: Technologies, Platforms and Paradigms for Real-Time Analytics of Data Streams

2021

Self Cite

View full text Add to dashboard Cite

Section: ) Data Stream Mining On the Go In Edge Environmentsmentioning

confidence: 99%

Section: ) Data Stream Mining On the Go In Edge Environmentsmentioning

confidence: 99%

Mapping the Big Data Landscape: Technologies, Platforms and Paradigms for Real-Time Analytics of Data Streams

2021

Self Cite

View full text Add to dashboard Cite

“…Distribution of machine learning methods running on smartphones, with the rise of Internet of Things, has been studied thoroughly in the area of Pocket Data Mining [14]. It is now evident that edge devices are complementary to cloud computing to scale out machine learning systems.…”

Section: Related Workmentioning

confidence: 99%

Edge Machine Learning: Enabling Smart Internet of Things Applications

Yazici

Basurra

Gaber

2018

BDCC

View full text Add to dashboard Cite

Machine learning has traditionally been solely performed on servers and high-performance machines. However, advances in chip technology have given us miniature libraries that fit in our pockets and mobile processors have vastly increased in capability narrowing the vast gap between the simple processors embedded in such things and their more complex cousins in personal computers. Thus, with the current advancement in these devices, in terms of processing power, energy storage and memory capacity, the opportunity has arisen to extract great value in having on-device machine learning for Internet of Things (IoT) devices. Implementing machine learning inference on edge devices has huge potential and is still in its early stages. However, it is already more powerful than most realise. In this paper, a step forward has been taken to understand the feasibility of running machine learning algorithms, both training and inference, on a Raspberry Pi, an embedded version of the Android operating system designed for IoT device development. Three different algorithms: Random Forests, Support Vector Machine (SVM) and Multi-Layer Perceptron, respectively, have been tested using ten diverse data sets on the Raspberry Pi to profile their performance in terms of speed (training and inference), accuracy, and power consumption. As a result of the conducted tests, the SVM algorithm proved to be slightly faster in inference and more efficient in power consumption, but the Random Forest algorithm exhibited the highest accuracy. In addition to the performance results, we will discuss their usability scenarios and the idea of implementing more complex and taxing algorithms such as Deep Learning on these small devices in more details.

show abstract

“…For example, the decision to operate in an island mode with the help of Photo Voltaic (PV) coupled with battery backup system or to operate in a grid connected mode is a decision that can be taken by the local MC. The local MC operates in a distributed manner and satisfies the idea of edge computing [33] where data processing is closer to the data source, thereby resulting in faster decision-making of resource utilization.…”

Section: Proposed Architecturementioning

confidence: 99%

Developing Self-Similar Hybrid Control Architecture Based on SGAM-Based Methodology for Distributed Microgrids

Gupta

Duchon

2018

Designs

View full text Add to dashboard Cite

Cyber-Physical Systems (CPS) are the complex systems that control and coordinate physical infrastructures, which may be geographically apart, via the use of Information and Communication Technology (ICT). One such application of CPS is smart microgrids. Microgrids comprise both power consuming and power producing infrastructure and are capable of operating in grid connected and disconnected modes. Due to the presence of heterogeneous smart devices communicating over multiple communication protocols in a distributed environment, a system architecture is required. The objective of this paper is to approach the microgrid architecture from the software and systems’ design perspective. The architecture should be flexible to support various multiple communication protocols and is able to integrate various hardware technologies. It should also be modular and scalable to support various functionalities such as island mode operations, energy efficient operations, energy trading, predictive maintenance, etc. These requirements are the basis for designing the software architecture for the smart microgrids that should be able to manage not only electrical but all energy related systems. In this work, we propose a distributed, hybrid control architecture suited for microgrid environments, where entities are geographically distant and need to operate in a cohesive manner. The proposed system architecture supports various design philosophies such as component-based design, hierarchical composition of components, peer-to-peer design, distributed decision-making and controlling as well as plug-and-play during runtime. A unique capability of the proposed system architecture is the self-similarity of the components for the distributed microgrids. The benefit of the approach is that it supports these design philosophies at all the levels in the hierarchy in contrast to a typical centralized architectures where decisions are taken only at the global level. The proposed architecture is applied to a real system of 13 residential buildings in a low-voltage distribution network. The required implementation and deployment details for monitoring and controlling 13 residential buildings are also discussed in this work.

show abstract

Pocket Data Mining

Cited by 25 publications

References 70 publications

Mapping the Big Data Landscape: Technologies, Platforms and Paradigms for Real-Time Analytics of Data Streams

Mapping the Big Data Landscape: Technologies, Platforms and Paradigms for Real-Time Analytics of Data Streams

Edge Machine Learning: Enabling Smart Internet of Things Applications

Developing Self-Similar Hybrid Control Architecture Based on SGAM-Based Methodology for Distributed Microgrids

Contact Info

Product

Resources

About