Forecasting framework for open access time series in energy

Barta, Gergo; Nagy, Gábor P.; Papp, Gyozo; Simon, Gábor

doi:10.1109/energycon.2016.7514015

Cited by 7 publications

(2 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The problem of time series prediction is a research topic for many years, which aims at predicting the value at a given time point. Time-series prediction problem is often addressed by using machine learning models, include Random Forest [15], GBRT [3], XGBoost [? ], and Long Short-Term Memory (LSTM) [2].…”

Section: B Flow Forecastmentioning

confidence: 99%

Which Channel to Ask My Question?: Personalized Customer Service Request Stream Routing Using Deep Reinforcement Learning

Liu

Long

et al. 2019

IEEE Access

View full text Add to dashboard Cite

Customer services are critical to all companies, as they may directly connect to the brand reputation. Due to a great number of customers, e-commerce companies often employ multiple communication channels to answer customers' questions, for example, chatbot and hotline. On one hand, each channel has limited capacity to respond to customers' requests, on the other hand, customers have different preferences over these channels. The current production systems are mainly built based on business rules, which merely considers tradeoffs between resources and customers' satisfaction. To achieve the optimal tradeoff between resources and customers' satisfaction, we propose a new framework based on deep reinforcement learning, which directly takes both resources and user model into account. In addition to the framework, we also propose a new deep-reinforcement-learning based routing method -double dueling deep Q-learning with prioritized experience replay (PER-DoDDQN). We evaluate our proposed framework and method using both synthetic and a real customer service log data from a large financial technology company. We show that our proposed deep-reinforcement-learning based framework is superior to the existing production system. Moreover, we also show our proposed PER-DoDDQN is better than all other deep Q-learning variants in practice, which provides a more optimal routing plan. These observations suggest that our proposed method can seek the trade-off where both channel resources and customers' satisfaction are optimal.INDEX TERMS Deep Reinforcement Learning, Personalized Customer Service, Time-series data processing

show abstract

Section: B Flow Forecastmentioning

confidence: 99%

Which Channel to Ask My Question?: Personalized Customer Service Request Stream Routing Using Deep Reinforcement Learning

Liu

Long

et al. 2019

IEEE Access

View full text Add to dashboard Cite

show abstract

“…In data science, our most recent research is about the renewable energy forecast by using distributed and privacy preserved machine-learning methods [180,181]. At end of the year, a successful R&D project was finished with OTP Bank, in which a new interpretable classification method was created and implemented to increase the efficiency of prediction.…”

Section: Data Science and Content Technologiesmentioning

confidence: 99%

BME VIK Annual Research Report on Electrical Engineering and Computer Science 2016

Charaf

Harsányi

Poppe

et al. 2017

Period. Polytech. Elec. Eng. Comp. Sci.

View full text Add to dashboard Cite

Preface Since being established in 1949, the Faculty of Electrical Engineering and Informatics (VIK) BME has played a flagship role in the development of electronics, IT and computer science in Hungary.We László Jakab (dean, BME VIK) János Levendovszky (vice-dean in charge of scientific affairs, BME VIK) are proud of combining engineering applications with sound scientific results, which is the assurance of high-level industrial collaboration leading to novel results and innovation. The various collaborations with our industrial partners has made it clear that the industry expects methods and results which make their industrial processes more effective and increase productivity and quality. Thus, participation in these collaborations give a competitive edge and ensure the continuous development of VIK. These factors have positioned our Department of Automation and Applied InformaticsDepartment of Automation and Applied Informatics (AUT) is one of the largest and dynamically developing departments at BME with diverse scope competences such as control theory, embedded systems, software modelling, applied software development, Internet of Things, power electronics, mechatronics, and many others. The department's main activities are education, research and development. Training versatile electrical and software engineers with solid practical knowledge is our top priority. Our profile also includes developing high quality software and hardware solutions for industry partners. All of our activities are backed by strong research background.In 2016, the department organized and hosted a number of domestic and international events. In cooperation with the industry, AUT has organized the IoT4U conference. It was dedicated to Internet of Things, cloud and data management as well as the integration of IoT into various formats. RobonAUT, the yearly domestic amateur robotics contest was organized by the department in February 2016. The yearly contest is open to teams of students, organized either in student projects within the department or independent research teams.The next sections summarize the key research results of the department in year 2016. Software Development Methods and Mobile PlatformsThe diversity of mobile platforms necessitates the development of the same mobile application for all major mobile platforms, what requires considerable development effort. Mobile application developers are multiplatform developers, but they still prioritize the platforms, therefore, not all platforms are equally important for them. Appropriate methods, processes and tools are required to support the development in order to achieve better productivity.The research on educational mobile applications that adapt to the mental state of the user required a method for estimating the difficulty of tasks found in the serious games. There are generated or random assignments where the effort required from the user is not known or cannot be estimated in advance.61 (2) IoT and Smart CityThe Internet of Things (IoT) is transforming the surroun...

show abstract

Review on probabilistic forecasting of photovoltaic power production and electricity consumption

Meer

Widén

Munkhammar

2018

Renewable and Sustainable Energy Reviews

316

140

View full text Add to dashboard Cite

Forecasting framework for open access time series in energy

Cited by 7 publications

References 7 publications

Which Channel to Ask My Question?: Personalized Customer Service Request Stream Routing Using Deep Reinforcement Learning

Which Channel to Ask My Question?: Personalized Customer Service Request Stream Routing Using Deep Reinforcement Learning

BME VIK Annual Research Report on Electrical Engineering and Computer Science 2016

Review on probabilistic forecasting of photovoltaic power production and electricity consumption

Contact Info

Product

Resources

About