The measurements of several sensors of multiple machines are collected in real-time at Alfa Laval’s IIoT platform. It is of great interest to make an accurate forecast for upcoming events, such as machine failure or sensor faults, which are critical enablers for predictive maintenance. However, some machines have little historical data, making forecasting challenging for them. Can we use data from other devices to make a forecast on a newly installed machine or improve the forecasting performance on different machines? This is the central question I try to answer in this research. However, another dimension of complexity makes the problem even more difficult. The devices have some differences. They can vary in version, application, and configuration. In addition, at each instant, the machine can have different modes and user-defined parameters that can be changed anytime by the operator or machine auto-control. All these factors make the forecasting problem extremely challenging, so it does not fit into a general computational framework. I propose a new multi-paradigm ensemble framework, aware of context and domain, that significantly improves the popular ensemble methods and non-ensemble state-of-the-art such as ARIMA and LSTM. The improvement mostly comes from adding a new forecasting paradigm based on multi-domain INN-DTW (one-nearest neighbor with dynamic time warping as a similarity measure) to the ensemble architecture, which creates a surprising harmony with ARIMA. INN-DTW is popular in time series classification, but its application is being investigated for the first time in a heterogeneous ensemble. The other novel component is a meta-selection mechanism for aggregating forecasts, which unexpectedly works better than the widely-used meta-learning approach.

Sensors are being used in many industrial applications for equipment health monitoring and anomaly detection. However, sometimes operation and maintenance of these sensors are costly. Thus companies are interested in reducing the number of required sensors as much as possible. The straightforward solution is to check the prediction power of sensors and eliminate those sensors with limited prediction capabilities. However, this is not an optimal solution because if we discard the identified sensors. As a result, their historical data also will not be utilized anymore. However, typically such historical data can help improve the remaining sensors' signal power, and abolishing them does not seem the right solution. Therefore, we propose the first data-driven approach based on tensor completion for re-utilizing data of removed sensors and the remaining sensors to create virtual sensors. We applied the proposed method on vibration sensors of high-speed separators, operating with five sensors. The producer company was interested in reducing the sensors to two. But with the aid of tensor completion-based virtual sensors, we show that we can safely keep only one sensor and use four virtual sensors that give almost equal detection power when we keep only two physical sensors.

MOTIVATION: Visualization of high-dimensional data is an important step in exploratory data analysis and knowledge discovery. However, it is challenging, because the interpretation is highly subjective. If we see dimensionality reduction (DR) techniques as the main tool for data visualization, they are like multiple cameras that look into the data from different perspectives or angles. We can hardly prescribe one single perspective for all datasets and problems. One snapshot of data cannot reveal all the relevant aspects of the data in higher dimensions. The reason is that each of these methods has its own specific strategy, normally based on well-established mathematical theories to obtain a low-dimensional projection of the data, which sometimes is totally different from the others. Therefore, relying only on one single projection can be risky, because it can close our eyes to important parts of the full knowledge space. RESULTS: We propose the first framework for multi-insight data visualization of multi-omics data. This approach, contrary to single-insight approaches, is able to uncover the majority of data features through multiple insights. The main idea behind the methodology is to combine several DR methods via tensor factorization and group the solutions into an optimal number of clusters (or insights). The experimental evaluation with low-dimensional synthetic data, simulated multi-omics data related to ovarian cancer, as well as real multi-omics data related to breast cancer show the competitive advantage over state-of-the-art methods. AVAILABILITY AND IMPLEMENTATION: https://folk.uio.no/hadift/MIV/ [user/pass via hadift@medisin. uio.no]. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Dimension reduction (DR) methods play an inevitable role in analyzing and visualizing high-dimensional multi-source data. In the recent decades many variants of these methods have been developed in various disciplines and domains. Due to the diversity and an ever-increasing number of developed techniques, choosing the right method for the given problem is a difficult task. In this study we benchmark 87 methods for integrative dimension reduction of mRNA expression and DNA methylation data, which is a common problem in biology and medicine. Our ranking is obtained based on four main factors: quality of dimension reduction (local, global, and local-global neighborhood preservation), clustering quality, speed and sensitivity to input parameters on multiple datasets generated by InterSIM (a semi-realistic multi-source data simulator in the genomics domain). The results are later validated on a real dataset for breast cancer through visual evaluation metrics such as co-ranking matrices, inspection of true cancer sub-types in two-dimensional projections, and LCMC curves. We also demonstrate the relationship between the methods via network inference. The findings in this study can be useful in algorithm selection and planning of experimental design in multi-source data analysis.

Link prediction is the task of social network analysis whose goal is to predict the links that will appear in the network in future instants. Among the link predictors exploiting the time evolution of the networks, we can find the tensor decomposition-based methods. A major limitation of these methods is the lack of appropriate approaches for estimating their parameters and initialization. In this paper, we address this problem by proposing a parameter setting method. Our proposed approach resorts to optimization techniques to drive the search for an adequate parameter and initialization choice.

Abstract Hotspot detection aims at identifying sub‐groups in the observations that are unexpected, with respect to some baseline information. For instance, in disease surveillance, the purpose is to detect sub‐regions in spatiotemporal space, where the count of reported diseases (e.g. cancer) is higher than expected, with respect to the population. The state‐of‐the‐art method for this kind of problem is the space–time scan statistics, which exhaustively search the whole space through a sliding window looking for significant spatiotemporal clusters. Space–time scan statistics makes some restrictive assumptions about the distribution of data, the shape of the hotspots and the quality of data, which can be unrealistic for some non‐traditional data sources. A novel methodology called EigenSpot is proposed where instead of an exhaustive search over the space, it tracks the changes in a space–time occurrences structure. The new approach does not only present much more computational efficiency but also makes no assumption about the data distribution, hotspot shape or the data quality. The principal idea is that with the joint combination of abnormal elements in the principal spatial and the temporal singular vectors, the location of hotspots in the spatiotemporal space can be approximated. The experimental evaluation, both on simulated and real data sets, reveals the effectiveness of the proposed method.

Syndromic surveillance systems continuously monitor multiple pre-diagnostic daily streams of indicators from different regions with the aim of early detection of disease outbreaks. The main objective of these systems is to detect outbreaks hours or days before the clinical and laboratory confirmati on. The type of data that is being generated via these systems is usually multivariate and seasonal with spatial and temporal dimensions. The algorithm What's Strange About Recent Events (WSARE) is the state-of-the-art method for such problems. It exhaustively searches for contrast sets in the multivariate data and signals an alarm when find statistically significant rules. This bottom-up approach presents a much lower detection delay comparing the existing top-down approaches. However, WSARE is very sensitive to the small-scale changes and subsequently comes with a relatively high rate of false alarms. We propose a new approach called EigenEvent that is neither fully top-down nor bottom-up. In this method, we instead of top-down or bottom-up search, track changes in data correlation structure via eigenspace techniques. This new methodology enables us to detect both overall changes (via eigenvalue) and dimension-level changes (via eigenvectors). Experimental results on hundred sets of benchmark data reveals that EigenEvent presents a better overall performance comparing state-of-the-art, in particular in terms of the false alarm rate.