Machine learning (ML) algorithms have gained prominence in time series prediction problems. Depending on the nature of the time series data, it can be difficult to build an accurate ML model with the proper structure and hyperparameters. In this study, we propose a predictive error compensation wavelet neural network model (PEC-WNN) for improving the prediction accuracy of chaotic and stochastic time series data. In the proposed model, an additional network is used for the prediction of the main network error to compensate the overall prediction error. The main network takes as inputs the time series data through moving frames in multiple-scales. The same structure and hyperparameter sets are applied for quite distinct four types of problems for verification of the robustness and accuracy of the proposed model. Specifically, the Mackey-Glass, Box-Jenkins, and Lorenz Attractor benchmark problems, as well as drought forecasting are used to characterize the performance of the model for chaotic and stochastic data cases. The results show that the PEC-WNN provides significantly more accurate predictions for all compared benchmark problems with respect to conventional machine learning and time series prediction methods without changing any hyperparameter or the structure. In addition, the time and space complexity of the PEC-WNN model is less than all other compared ML methods, including long short-term memory (LSTM) and convolutional neural networks (CNNs).

In this study, the phenological behavior, and energy balance of plants are used as a sensory mechanism for root-zone soil moisture monitoring using both in-situ and satellite remote sensing data. The commonly used in-situ measurements are not feasible for large-scale agricultural areas. Local direct soil moisture measurements cannot be reliably interpolated owing to the high spatial variability of soil structure and the vegetative content. Remote sensing methods are affected by vegetation coverage and density owing to penetration and backscattering characteristics. We propose a root-zone soil moisture estimation method, focused on a context-aware data clustering process, which can be applied prior to any statistical analysis, for empirical evaluation of data. In this aspect, the crops phenological stages and soilair temperature differences are defined as the two contexts for clustering. Parameters such as canopyair temperature difference, land surface temperature, and solar radiation with respect to plant energy and water processes are defined for the analysis. The proposed model is utilized using piecewise linear regression of data obtained from 16 rainfed wheat parcels distributed across Turkey, under different climatic and topographic conditions. It is shown that the proposed context-aware data clustering process enables the nonlinear agricultural system to be analyzed linearly. The correlation value of the whole season increased from 21% to correlation values changing between 78% and 95% for different clusters. The outliers became relevant and the parameters became significant after the proposed context-aware data clustering.

Efficiently transferring information, which is of major importance for almost all engineering systems, requires a data representative appropriate to maintain the density and consistency of the information dispatched. As such a machine-readable representation of data, barcodes are one of the most recognized and powerful instruments of the purpose in every domain. Likewise, from the perspective of engineering, colors are excellent sources of information exchange, since the transmission of color connotes the conveyance of its entire scalar attributes in the same spatial channel component. Exerting colors on barcodes as an effective way of bursting the data conveyance capacity has been an active area of research for over 50 years. Significant progress has been achieved through efforts in this regard. It is also envisagable that the evident evolution in related technologies exoterically empowers the enhancement of color barcode capabilities as capacity and reliability, thereby further encouraging prospected research in this direction. Herein, a comprehensive survey of the studies on this main area of interest is presented. To help better acquainted with the field, also a taxonomy of the peculiar interference sources and distortion effects is provided, besides, the 3D barcoding process itself and the research areas are described. Most of the relevant works from debut to the present are broadly examined. Rather than presenting a timeline, studies that pertain to similar issues are addressed together. Amongst those related, premising or pivotal ones are preferably cited as far as feasible. Moreover, all perused works are analyzed by the research areas and the results are presented. Also, the issues relatively more prominent as affecting the performance of the whole process are specified. In the conclusion, some of the research subjects that appear open, scarce, or require further elaboration were remarked on as well. It is anticipated this study to contribute to the efforts toward leveraging color in barcodes.

Inefficient utilization of the frequency spectrum due to conventional regulatory limitations and physical performance limiting factors, mainly the Signal to Noise Ratio (SNR), are prominent restrictions in digital wireless communication. Pattern Based Communication System (PBCS) is an adaptive and perceptual communication method based on a Cognitive Radio (CR) approach. It intends an SNR oriented cognition mechanism in the physical layer for improvement of Link Spectral Efficiency (LSE). The key to this system is construction of optimal communication signals, which consist of encoded data in different pattern forms (waveforms) depending on spectral availabilities. The signals distorted in the communication medium are recovered according to the pre-trained pattern glossary by the perceptual receiver. In this study, we have shown that it is possible to improve the bandwidth efficiency when largely uncorrelated signal patterns are chosen in order to form a glossary that represents symbols for different length data groups and the information can be recovered by the Artificial Neural Network (ANN) in the receiver site.

Multivariable machine learning (ML) models are increasingly used for time series predictions. However, avoiding the overfitting and underfitting in ML-based time series prediction requires special consideration depending on the size and characteristics of the available training dataset. Predictive error compensating wavelet neural network (PEC-WNN) improves the time series prediction accuracy by enhancing the orthogonal features within a data fusion scheme. In this study, time series prediction performance of the PEC-WNNs have been evaluated on two different problems in comparison to conventional machine learning methods including the long short-term memory (LSTM) network. The results have shown that PECNET provides significantly more accurate predictions. RMSPE error is reduced by more than 60% with respect to other compared ML methods for Lorenz Attractor and wind speed prediction problems.