3 December 2021 in IISE Transactions
In recent years, information available from multiple data modalities has become increasingly common for industrial engineering and operations research applications. There have been a number of research works combining these data in unsupervised, supervised, and semi-supervised fashions that have addressed various issues of combining heterogeneous data, as well as several existing open challenges that remain to be addressed. In this review paper, we provide an overview of some methods for the fusion of multimodal data. We provide detailed real-world examples in manufacturing and medicine, introduce early, late, and intermediate fusion, as well as discuss several approaches under decomposition-based and neural network fusion paradigms. We summarize the capabilities and limitations of these methods and conclude the review article by discussing the existing challenges and potential research opportunities.
14 January 2020 in Technometrics
In recent years, measurement or collection of heterogeneous sets of data such as those containing scalars, waveform signals, images, and even structured point clouds, has become more common. Statistical models based on such heterogeneous sets of data that represent the behavior of an underlying system can be used in the monitoring, control, and optimization of the system. Unfortunately, available methods mainly focus on the scalars and profiles and do not provide a general framework for integrating different sources of data to construct a model. This paper addresses the problem of estimating a process output, measured by a scalar, curve, image, or structured point cloud by a set of heterogeneous process variables such as scalar process setting, profile sensor readings, and images. We introduce a general multiple tensor-on-tensor regression (MTOT) approach in which each set of input data (predictor) and output measurements are represented by tensors. We formulate a linear regression model between the input and output tensors and estimate the parameters by minimizing a least square loss function. In order to avoid overfitting and reduce the number of parameters to be estimated, we decompose the model parameters using several basis matrices that span the input and output spaces, and provide efficient optimization algorithms for learning the basis and coefficients. Through several simulation and case studies, we evaluate the performance of the proposed method. The results reveal the advantage of the proposed method over some benchmarks in the literature in terms of the mean square prediction error. supplementary materials for your article are available online.
1 December 2012 in Transportation Research Part C: Emerging Technologies
1 January 2013 in Mathematical Biosciences and Engineering
SDSiamak Mahmoudian Dehkordi
Siamak Mahmoudian Dehkordi
In standard chemotherapy protocols, drugs are given at maximum tolerated doses(MTD) with rest periods in between. In this paper, we briefly discuss therationale behind this therapy approach and, using as examplemulti-drug cancer chemotherapy with a cytotoxic and cytostatic agent, show thatthese types of protocols are optimal in the sense of minimizing a weightedaverage of the number of tumor cells (taken both at the end of therapy and atintermediate times) and the total dose given if it is assumed that the tumorconsists of a homogeneous population of chemotherapeutically sensitive cells.A $2$-compartment linear model is used to model the pharmacokinetic equations for the drugs.
2 October 2018 in Journal of Quality Technology
While anomaly detection in static networks has been extensively studied, only recently have researchers focused on dynamic networks. This trend is mainly due to the capacity of dynamic networks to represent complex physical, biological, cyber, and social systems. This article proposes a new methodology for modeling and monitoring dynamic attributed networks for quick detection of temporal changes in network structures. In this methodology, the generalized linear model (GLM) is used to model static attributed networks. This model is then combined with a state transition equation to capture the dynamic behavior of the system. Extended Kalman filter (EKF) is used as an online, recursive inference procedure to predict and update network parameters over time. In order to detect changes in the underlying mechanism of edge formation, prediction residuals are monitored through an exponentially weighted moving average (EWMA) control chart. The proposed modeling and monitoring procedure is examined through simulations for attributed binary and weighted networks. Email communication data from the Enron corporation is used as a case study to show how the method can be applied in real-world problems.
1 February 2019 in Acta Materialia