Reconstruction of time series with missing value using 2D representation-based denoising autoencoder

Time series analysis is a key technology for medical diagnosis,weather forecasting and financial prediction systems.However,missing data frequently occur during data recording,posing a great challenge to data mining tasks.In this study,we propose a novel time series data representation-based denoising autoencoder(DAE)for the reconstruction of missing values.Two data representation methods,namely,recurrence plot(RP)and Gramian angular field(GAF),are used to transform the raw time series to a 2D matrix for establishing the temporal correlations between different time intervals and extracting the structural patterns from the time series.Then an improved DAE is proposed to reconstruct the missing values from the 2D representation of time series.A comprehensive comparison is conducted amongst the different representations on standard datasets.Results show that the 2D representations have a lower reconstruction error than the raw time series,and the RP representation provides the best outcome.This work provides useful insights into the better reconstruction of missing values in time series analysis to considerably improve the reliability of timevarying system.

Transition Dipole, Charge Transfer, and Electron-hole Coherence in Two-photon Absorption： Visualizations with Two Dimensional Site and Three Dimensional Cube Representations

The developed visualization methods of two dimensional （2D） site and three dimensional （3D） cube representations have been performed to show the orientation of transition dipole, charge transfer, and electron-hole coherence in two-photon absorption （TPA）. The 3D cube representations of transition density can reveal visually the orientation and strength of transition dipole moment, and charge different density show the orientation of charge transfer in TPA. The 2D site representation can reveal visually the electron-hole coherence in TPA. The combination of 2D site and 3D cube representations provide clearly inspect into the charge transfer process and the contribution of excited molecular segments for TPA.

A new 2D graphical representation of protein sequence and its application

Graphical representation is a very efficient tool for visual analysis of protein sequences. In this paper, a novel 2D graphical representation scheme is proposed on the basis of a newly introduced concept, named characteristic model of the protein sequences. After obtaining the 2D graphics of protein sequences, two numerical characterizations of them is designed as descriptors to analyze the nine DN5 protein sequences, simulation and analysis results show that, comparing with existing methods, our method is not only visible, intuitional, and simple, but also has no circuit or degeneracy, and even more important, since the storage space required by our method is constant and has nothing to do with the length of protein sequences, then it can keep excellent visual inspection for long protein sequences.

A ubiquitous knowledgeable data representation model(UKRM) for three-dimensional geographic information systems(3D GIS)

In the face of complicated, diversified three-dimensional world, the existing 3D GIS data models suffer from certain issues such as data incompatibility, insufficiency in data representation and representation types, among others. It is often hard to meet the requirements of multiple application purposes(users) related to GIS spatial data management and data query and analysis, especially in the case of massive spatial objects. In this study, according to the habits of human thinking and recognition, discrete expressions(such as discrete curved surface(DCS), and discrete body(DB)) were integrated and two novel representation types(including function structure and mapping structure) were put forward. A flexible and extensible ubiquitous knowledgeable data representation model(UKRM) was then constructed, in which structurally heterogeneous multiple expressions(including boundary representation(B-rep), constructive solid geometry(CSG), functional/parameter representation, etc.) were normalized. GIS's ability in representing the massive, complicated and diversified 3D world was thus greatly enhanced. In addition, data reuse was realized, and the bridge linking static GIS to dynamic GIS was built up. Primary experimental results illustrated that UKRM was overwhelmingly superior to the current data models(e.g. IFC, City GML) in describing both regular and irregular spatial objects.