scispace - formally typeset
Search or ask a question
Author

Zhao Zhikai

Bio: Zhao Zhikai is an academic researcher from China University of Mining and Technology. The author has contributed to research in topics: Interface (computing) & Deep learning. The author has an hindex of 2, co-authored 2 publications receiving 15 citations.

Papers
More filters
Patent
11 Nov 2015
TL;DR: In this paper, the authors proposed a method for reducing the power consumption of the Wi-Fi network interface card based on data importance, characterized in that the method includes the following steps: (1) assessing the application which requests the network data, and determining a delayed sensitivity level of the application; (2) based on the delayed sensitivity levels of application, via a corresponding wifi power consumption optimization scheme, configuring a wifi network interfaces card for controlling the transmission of network data.
Abstract: The invention discloses a method for lowering wifi power consumption based on data importance, characterized in that the method includes the following steps: (1) based on the sensitivity to network data of an application which requests the network data, assessing the application which requests the network data, and determining a delayed sensitivity level of the application; (2) based on the delayed sensitivity level of the application, via a corresponding wifi power consumption optimization scheme, configuring a wifi network interface card for controlling the transmission of the network data. The method herein can reduce tail part energy consumption of the network interface card. The delay in waking up the network interface card can increase the time when the network interface card is in sleep mode.

12 citations

Proceedings ArticleDOI
01 Jan 2015
TL;DR: Results show that the features obtained by deep learning are more descriptive for predicting terminal replacement behavior than three others models based on 1 Nearest Neighbors, Support Vector Machines and Neural Network.
Abstract: To help telecommunications operators accurately predict the terminal replacement behavior, and improve the success rate of marketing and the accuracy of resources devoting, huge user consumption data are used to build Deep Belief Network. The deep features that characterize the terminal replacement behavior are learned, through which a terminal replacement prediction model is conducted. Experiments are carried out on real data set, and the prediction accuracy is over 82%. It is better than three others models based on 1 Nearest Neighbors, Support Vector Machines and Neural Network. The experiments results show that the features obtained by deep learning are more descriptive for predicting terminal replacement behavior.

4 citations


Cited by
More filters
Journal ArticleDOI
TL;DR: A thorough investigation of deep learning in its applications and mechanisms is sought, as a categorical collection of state of the art in deep learning research, to provide a broad reference for those seeking a primer on deep learning and its various implementations, platforms, algorithms, and uses in a variety of smart-world systems.
Abstract: Deep learning has exploded in the public consciousness, primarily as predictive and analytical products suffuse our world, in the form of numerous human-centered smart-world systems, including targeted advertisements, natural language assistants and interpreters, and prototype self-driving vehicle systems. Yet to most, the underlying mechanisms that enable such human-centered smart products remain obscure. In contrast, researchers across disciplines have been incorporating deep learning into their research to solve problems that could not have been approached before. In this paper, we seek to provide a thorough investigation of deep learning in its applications and mechanisms. Specifically, as a categorical collection of state of the art in deep learning research, we hope to provide a broad reference for those seeking a primer on deep learning and its various implementations, platforms, algorithms, and uses in a variety of smart-world systems. Furthermore, we hope to outline recent key advancements in the technology, and provide insight into areas, in which deep learning can improve investigation, as well as highlight new areas of research that have yet to see the application of deep learning, but could nonetheless benefit immensely. We hope this survey provides a valuable reference for new deep learning practitioners, as well as those seeking to innovate in the application of deep learning.

411 citations

Patent
17 Aug 2016
TL;DR: In this paper, a control method, a control device and a control system for an IP (Internet Protocol) camera is presented, which is capable of preventing the IP camera from being in the working state all the time.
Abstract: The invention provides a control method, a control device and a control system for an IP (Internet Protocol) Camera. The method comprises the following steps: the IP camera waits for receiving a standby command, wherein the IP camera is currently in a working state; if receiving the standby command transmitted by a server, the IP camera is switched from the working state to a standby state. The control method, the control device and the control system which are provided by the invention are capable of preventing the IP camera from being in the working state all the time, so that the power consumption of the IP camera is reduced.

19 citations

Journal ArticleDOI
TL;DR: A model centered on Deep Learning (DL) for predicting complications of Type 2 Diabetes Mellitus is proposed, which follows data collection, pre-training, feature extraction, Deep Belief Network (DBN), validation process, and classification steps for predicting diabetic complications.
Abstract: The revolution in digitization makes the health care sector as a prime source of big data. The analysis of these data could be a great supporting source for deriving new insights, which increases the care and awareness about health. Diabetes together with its complications has been recognized worldwide as a chief public health threat. Predicting diabetic complications is considered as a highly effectual technique for augmenting the survival rate of diabetic patients. While many studies currently use medical images and structured medical records, very limited efforts have been dedicated for applying Data Mining (DM) techniques for unstructured textual medical records, for instance, admission and discharge records. Many DM techniques have been generated for envisaging diabetic complications. But in existing methods, the classification as well as prediction accuracy is not so high. So this paper proposes a model centered on Deep Learning (DL) for predicting complications of Type 2 Diabetes Mellitus. The proposed model follows data collection, pre-training, feature extraction, Deep Belief Network (DBN), validation process, and classification steps for predicting diabetic complications. Finally, the performances proffered by the proposed DL based Big Medical Data Analytics model using DBN as well as the prevailing techniques are contrasted with respect to Precision, accuracy, and Recall. The Training, as well as the Testing process, delineates the pervasiveness of risk with an accuracy of 81.20%. This realistic prediction model will be very much useful for effectively managing diabetes.

12 citations

Patent
08 Aug 2017
TL;DR: In this paper, an information processing method, device and a mobile terminal are used to monitor the power consumption type of the standby power consumption of the mobile terminal, and the state information of the Wi-Fi connection changes.
Abstract: The invention relates to an information processing method, device and a mobile terminal. The method comprises the following steps: receiving a screen-off instruction from the mobile terminal, and performing zero clearing for the statistics number of state information of wireless fidelity Wi-Fi connection; if detecting that the state information of the Wi-Fi connection changes, updating the statistics number of the state information of the Wi-Fi connection; and receiving a screen-on instruction from the mobile terminal, stopping detecting the state information of the Wi-Fi connection. By adopting the method disclosed by the invention, the power consumption type of the standby power consumption of the mobile terminal can be monitored.

12 citations

Patent
31 May 2017
TL;DR: In this paper, a smart phone energy consumption optimization method based on a set optimization algorithm is proposed. And the method mainly comprises three assemblies: a monitoring assembly, a mining assembly and a scheduling assembly.
Abstract: The invention discloses a smart phone energy consumption optimization method based on a set optimization algorithm. The method is realized by middleware software of a cross application. The method mainly comprises three assemblies: a monitoring assembly, a mining assembly and a scheduling assembly. Network activity data used by a user when a screen is closed is collected; the importance of an application for the user is predicted through utilization of a decision-making tree; the obtained value of the importance is introduced into the set optimization algorithm; a network request of the application is constrained; network activities after the screen is closed are reduced; and the battery loss resulting from the network activities is reduced. On the premise of ensuring the user experience, the duration and energy saving performance of a mobile phone are maximized.

8 citations