Time-Series Data Prediction Based on Trending Structure Sequence and Rough Set

Adding the dimension of time to databases produces time series databases (TSDB) and introduces new aspects and difficulties to data mining and knowledge discovery. In this correspondence, we introduce a general methodology for knowledge discovery in TSDB. The process of knowledge discovery in TSDR includes cleaning and filtering of time series data, identifying the most important predicting attributes, and extracting a set of association rules that can be used to predict the time series behavior in the future. Our method is based on signal processing techniques and the information-theoretic fuzzy approach to knowledge discovery. The computational theory of perception (CTP) is used to reduce the set of extracted rules by fuzzification and aggregation. We demonstrate our approach on two types of time series: stock-market data and weather data.

When the time dimension is added to datasets, time series data are obtained. Extracting knowledge from time series data requires special attention to the timing aspects of the data. An interesting activity in the field of knowledge discovery from time series data is predicting the timing of upcoming events. In this paper we present a method for mining fuzzy knowledge from time series data. In contrast to traditional time series analysis methods which largely focus on global models, our method is about the discovery of local patterns in time series. The extracted knowledge will be in the form of fuzzy association rules and it aims at predicting the approximate timing of upcoming events. The proposed method includes cleaning and filtering of time series data, segmenting time series, extracting important features for prediction, further cleaning on feature values, fuzzifying feature values, extracting fuzzy association rules, and pruning the discovered rules. We will show the efficiency of our approach on a stock market dataset.Keywordsknowledge discoverytime seriesfuzzy association Rulespreprocessingpost-processing

Time series similarity search is an important aspect of knowledge discovery in time series, and the index of data is the key to efficiency of similarity search in time series. In order to improve the efficiency of data query, a new similarity search method is proposed for hydrological time series using Lucene function mechanism. In the initial stage, build index for data with Lucene, and then carry out wavelet transform and symbolic on data, based on the candidate set with similar symbol selected from the sequence, and utilize range search function provided by Lucene rapidly search the sub sequence from candidate set, and use DTW (dynamic time warping) algorithm to calculate the matched sequence closest to search sequence. Finally, the validity and rationality of the method are verified by using the data of a hydrological station.

Daily time series of microwave radiometer data obtained in one-orbit direction are full of observation gaps due to satellite configuration and errors from spatial sampling. Such time series carry information about the surface signal including surface emittance and vegetation attenuation, and the atmospheric signal including atmosphere emittance and atmospheric attenuation. To extract the surface signal from this noisy time series, the Time Series Analysis Procedure (TSAP) was developed, based on the properties of the Discrete Fourier Transform (DFT). TSAP includes two stages: (1) identify the spectral features of observation gaps and errors and remove them with a modified boxcar filter; and (2) identify the spectral features of the surface signal and reconstruct it with the Harmonic Analysis of Time Series (HANTS) algorithm. Polarization Difference Brightness Temperature (PDBT) at 37 GHz data were used to illustrate the problems, to explain the implementation of TSAP and to validate this method, due to the PDBT sensitivity to the water content both at the land surface and in the atmosphere. We carried out a case study on a limited heterogeneous crop land and lake area, where the power spectrum of the PDBT time series showed that the harmonic components associated with observation gaps and errors have periods ≤8 days. After applying the modified boxcar filter with a length of 10 days, the RMSD between raw and filtered time series was above 11 K, mainly related to the power reduction in the frequency range associated with observation gaps and errors. Noise reduction is beneficial when applying PDBT observations to monitor wet areas and open water, since the PDBT range between dryland and open water is about 20 K. The spectral features of the atmospheric signal can be revealed by time series analysis of rain-gauge data, since the PDBT at 37 GHz is mainly attenuated by hydrometeors that yield precipitation. Thus, the spectral features of the surface signal were identified in the PDBT time series with the help of the rain-gauge data. HANTS reconstructed the upper envelope of the signal, i.e., correcting for atmospheric influence, while retaining the spectral features of the surface signal. To evaluate the impact of TSAP on retrieval accuracy, the fraction of Water Saturated Surface (WSS) in the region of Poyang Lake was retrieved with 37 GHz observations. The retrievals were evaluated against estimations of the lake area obtained with MODerate-resolution Imaging Spectroradiometer (MODIS) and Advanced Synthetic Aperture Radar (ASAR) data. The Relative RMSE on WSS was 39.5% with unfiltered data and 23% after applying TSAP, i.e., using the estimated surface signal only.

The LNCS journal Transactions on Rough Sets is devoted to the entire spectrum of rough sets related issues, from logical and mathematical foundations, through all aspects of rough set theory and its applications, such as data mining, knowledge discovery, and intelligent information processing, to relations between rough sets and other approaches to uncertainty, vagueness, and incompleteness, such as fuzzy sets and theory of evidence. Volume XVI includes extensions of papers from the Rough Sets and Knowledge Technology Conference which was held in Banff, Canada, in October 2011. In addition this book contains a long paper based on a PhD thesis. The papers cover both theory and applications of rough, fuzzy and near sets. They offer a continuation of a number of research streams which have grown out of the seminal work by Zdzislaw Pawlak during the first decade of the 21st century.

Binary information table, multi-valued information table and set-valued information table are three kinds of information systems often encountered in information processing. For any information system, we can often induce different information granular structures, and then construct the corresponding rough set models. Generally speaking, for the same information system, three models of Pawlak rough set, covering rough set and multi-granulation rough set can be induced according to different rules. These three kinds of rough set models are effective tools for data mining and information processing. This paper studies the relationship among Pawlak rough set, covering rough set and multi-granularity rough set induced in binary information table, multi-valued information table and set-valued information table, and obtains many important conclusions. The research content of this paper effectively connects the theories, methods and applications of Pawlak rough set, covering rough set and multi-granularity rough set, which not only enriches the rough set theory, but also expands the application prospect of rough set.

Assessing reliability of Big Data Knowledge Discovery process

The export value of East Kalimantan Province has big data conditions with time series and multivariable data types. Cluster analysis can be applied to time series data, where there are different procedures and grouping algorithms compared to grouping cross section data. Algorithms and procedures in the cluster formation process are done differently, because time series data is a series of observational data that occur based on a time index in sequence with a fixed time interval. The purpose of this research is to obtain the best similarity measurement using the cophenetic correlation coefficient and get the optimal c-value using the silhouete coefficient. In this study, the grouping algorithm used is a single linkage with four measurements of similarity, namely the Pearson correlation distance, euclidean, dynamic time warping and autocorrelation based distance. The sample in this study is the data on the export value of oil and non-oil commodities in East Kalimantan Province from January 2000 to December 2016 consisting of 10 variables. Based on the results of the analysis, the distance of the best similarity measurement in clustering the export value of oil and non-oil commodities in East Kalimantan Province is the dynamic time warping distance with the optimal c-value of 3 clusters.

In order to provide for the nature of the various deformations effectively and discover the deformation regularity and trend of deformation body, people propose an indirect deformation prediction method. It applies the segmented model of multidimensional time series to present an inertial measurement parameters. The method applies the MIMU measurement unit to realize data collection. Through the analysis of filter compensation, it conducts the deformation prediction. MIMU signal changes on time rate and has high noise. According to the analysis of error sources of MIMU signal, we adopt the method of time series analysis to data preprocessing and establish the AR model of the time series. By using AR-Kalman filter algorithm, we take care of random error processing. Due to deformation occurs more slowly. It’s difficult for the deformation prediction of the time series data after filtering and MIMU directly calculats posture. So this paper through the calculation of similarity of each period of time series predicts the deformation tendency of deformation. Through the acquisition of the results of the actual data of the bridge deformation, this paper introduces the deformation prediction method is correct and effective, and it can be implemented to the deformation forecast in practice.

Financial firms commonly process and store billions of time-series data, generated continuously and at a high frequency. To support efficient data storage and retrieval, specialized time-series databases and systems have emerged. These databases support indexing and querying of time-series by a constrained Structured Query Language(SQL)-like format to enable queries like "Stocks with monthly price returns greater than 5%", and expressed in rigid formats. However, such queries do not capture the intrinsic complexity of high dimensional time-series data, which can often be better described by images or language (e.g., "A stock in low volatility regime"). Moreover, the required storage, computational time, and retrieval complexity to search in the time-series space are often non-trivial. In this paper, we propose and demonstrate a framework to store multi-modal data for financial time-series in a lower-dimensional latent space using deep encoders, such that the latent space projections capture not only the time series trends but also other desirable information or properties of the financial time-series data (such as price volatility). Moreover, our approach allows user-friendly query interfaces, enabling natural language text or sketches of time-series, for which we have developed intuitive interfaces. We demonstrate the advantages of our method in terms of computational efficiency and accuracy on real historical data as well as synthetic data, and highlight the utility of latent-space projections in the storage and retrieval of financial time-series data with intuitive query modalities.

Advances in forecasting with neural networks? Empirical evidence from the NN3 competition on time series prediction

Time series (TS) data is ubiquitous in various fields such as finance, economics, meteorology, and engineering. The analysis of TS data aims to understand the underlying patterns, make predictions, and inform decision-making. One of the fundamental techniques in TS analysis is decomposition, which breaks down a TS into its constituent components: trend, seasonality, and residuals. This chapter provides a comprehensive overview of TS decomposition, breaking down data into trend, seasonality, and residuals. It covers classical methods, such as additive and multiplicative models, advanced techniques like X-12-ARIMA and Seasonal-Trend decomposition using LOESS (STL), and recent approaches, including machine learning (ML) based decompositions. Practical applications in agriculture, meteorology, and economics, along with challenges like non-stationarity and nonlinear behavior, are discussed. The chapter offers guidelines for selecting appropriate methods and includes case studies for real-world insights. It is a valuable resource for researchers, data scientists, and professionals analyzing complex TS data.

It has long been recognized that age-structure data contain useful information for assessing the status and dynamics of wildlife populations. For example, age-specific survival rates can be estimated with just a single sample from the age distribution of a stable, stationary population. For a population that is not stable, age-specific survival rates can be estimated using techniques such as inverse methods that combine time series of age-structure data with other demographic data. However, estimation of survival rates using these methods typically requires numerical optimization, a relatively long time series of data, and smoothing or other constraints to provide useful estimates. We developed general models for possibly unstable populations that combine time series of age-structure data with other demographic data to provide explicit maximum likelihood estimators of age-specific survival rates with as few as two years of data. As an example, we applied these methods to estimate survival rates for female bison (Bison bison) in Yellowstone National Park, USA. This approach provides a simple tool for monitoring survival rates based on age-structure data.

The increased availability of time series datasets prompts the development of new tools and methods that allow machine learning classifiers to better cope with time series data. Time series data are usually characterized by a high space dimensionality and a very strong correlation among features. This special nature makes the development of effective time series classifiers a challenging task. This work proposes and analyzes methods combining spectral decomposition and feature selection for time series classification problems and compares them against methods that work with original time series and time-dependent features. Briefly, our approach first applies discrete cosine transform (DCT) or discrete wavelet transform (DWT) on time series data. Then, it performs supervised feature selection/reduction by selecting only the most discriminative set of coefficients to represent the data. Experimental evaluations, carried out on multiple datasets, demonstrate the benefits of our approach in learning efficient and accurate time series classifiers.

Notice of Violation of IEEE Publication Principles<br><br>"Knowledge Discovery from Virtual Enterprise Model Based on Semantic Annotation,"<br>by Chengzhu Sun, Xiaofei Xu, Xiangyang Li, Shengchun Deng,<br>in the Proceedings of the Fifth International Conference on Fuzzy Systems and Knowledge Discovery, 2008. FSKD '08, vol.5, pp.546-551, Oct. 2008<br><br>After careful and considered review of the content and authorship of this paper by a duly constituted expert committee, this paper has been found to be in violation of IEEE's Publication Principles.<br><br>This paper contains significant portions of original text from the paper cited below. The original text was copied without attribution (including appropriate references to the original author(s) and/or paper title) and without permission.<br><br>"A Controlled Language for Semantic Annotation and Interoperability in e-Business Applications"<br>by M.Missikoff, F.Schiappelli , F.Taglino<br>in the Proceedings of the Semantic Integration Workshop, collocated with the Second International Semantic Web Conference (ISWC-03), October 2003<br><br> <br/> Discovering knowledge from virtual enterprise model is becoming increasingly important, as numerical models established for virtual enterprise are difficult to support interoperability of virtual enterprise. To solve this problem, a knowledge discovery method based on semantic annotation is put forward in this paper. A graphic ontology representation schema for virtual enterprise model is described, which is called ontology structure graph (OSG). Based on applications of semantic annotation, the process of knowledge discovery in the virtual enterprise model base is given and activities such as model selection, semantic annotation, data transformation, knowledge extraction and ontology interoperation in knowledge discovery process are illustrated in detail. Then, several critical issues influencing knowledge discovery are explained, including the organization of domain vocabulary, the definition of semantic annotation rules and semantic affinity function, the formulation of reference ontology. Finally, an instance is given to demonstrate the knowledge discovery method and results of knowledge discovery are presented.