scholarly journals Machine Learning Assisted Solutions of Mixed Integer MPC on Embedded Platforms

2020 ◽  
Vol 53 (2) ◽  
pp. 5195-5200
Author(s):  
Yannik Löhr ◽  
Martin Klaučo ◽  
Miroslav Fikar ◽  
Martin Mönnigmann
Keyword(s):  
Machine Learning ◽  
Mixed Integer ◽  
Embedded Platforms

scholarly journals Soft Computing Methodology to Optimize the Integrated Dynamic Models of Cellular Manufacturing Systems in a Robust Environment

2021 ◽  
Vol 2021 ◽  
pp. 1-13
Author(s):  
Amir-Mohammad Golmohammadi ◽  
Hasan Rasay ◽  
Zaynab Akhoundpour Amiri ◽  
Maryam Solgi ◽  
Negar Balajeh
Keyword(s):  
Machine Learning ◽  
Programming Model ◽  
Facility Layout ◽  
Cell Formation ◽  
Layout Design ◽  
Metaheuristic Algorithms ◽  
Sensitivity Analyses ◽  
Mixed Integer ◽  
Facility Layout Design ◽  
Proposed Model

Machine learning, neural networks, and metaheuristic algorithms are relatively new subjects, closely related to each other: learning is somehow an intrinsic part of all of them. On the other hand, cell formation (CF) and facility layout design are the two fundamental steps in the CMS implementation. To get a successful CMS design, addressing the interrelated decisions simultaneously is important. In this article, a new nonlinear mixed-integer programming model is presented which comprehensively considers solving the integrated dynamic cell formation and inter/intracell layouts in continuous space. In the proposed model, cells are configured in flexible shapes during the planning horizon considering cell capacity in each period. This study considers the exact information about facility layout design and material handling cost. The proposed model is an NP-hard mixed-integer nonlinear programming model. To optimize the proposed problem, first, three metaheuristic algorithms, that is, Genetic Algorithm (GA), Keshtel Algorithm (KA), and Red Deer Algorithm (RDA), are employed. Then, to further improve the quality of the solutions, using machine learning approaches and combining the results of the aforementioned algorithms, a new metaheuristic algorithm is proposed. Numerical examples, sensitivity analyses, and comparisons of the performances of the algorithms are conducted.

scholarly journals Matheuristics and Column Generation for a Basic Technician Routing Problem

Algorithms ◽  
2021 ◽  
Vol 14 (11) ◽  
pp. 313
Author(s):  
Nicolas Dupin ◽  
Rémi Parize ◽  
El-Ghazali Talbi
Keyword(s):  
Machine Learning ◽  
Column Generation ◽  
Time Window ◽  
Time Windows ◽  
Machine Learning Techniques ◽  
Mixed Integer ◽  
Routing Problems ◽  
Routing Problem ◽  
Feasible Solutions ◽  
Time Window Constraints

This paper considers a variant of the Vehicle Routing Problem with Time Windows, with site dependencies, multiple depots and outsourcing costs. This problem is the basis for many technician routing problems. Having both site-dependency and time window constraints lresults in difficulties in finding feasible solutions and induces highly constrained instances. Matheuristics based on Mixed Integer Linear Programming compact formulations are firstly designed. Column Generation matheuristics are then described by using previous matheuristics and machine learning techniques to stabilize and speed up the convergence of the Column Generation algorithm. The computational experiments are analyzed on public instances with graduated difficulties in order to analyze the accuracy of algorithms for ensuring feasibility and the quality of solutions for weakly to highly constrained instances. The results emphasize the interest of the multiple types of hybridization between mathematical programming, machine learning and heuristics inside the Column Generation framework. This work offers perspectives for many extensions of technician routing problems.

scholarly journals Machine-Learning Methods to Select Potential Depot Locations for the Supply Chain of Biomass Co-Firing

Energies ◽  
2020 ◽  
Vol 13 (24) ◽  
pp. 6554
Author(s):  
Diana Goettsch ◽  
Krystel K. Castillo-Villar ◽  
Maria Aranguren
Keyword(s):  
United States ◽  
Machine Learning ◽  
Supply Chain ◽  
Ghg Emissions ◽  
The United States ◽  
Mixed Integer ◽  
Total Cost ◽  
Computational Burden ◽  
Milp Model

Coal is the second-largest source for electricity generation in the United States. However, the burning of coal produces dangerous gas emissions, such as carbon dioxide and Green House Gas (GHG) emissions. One alternative to decrease these emissions is biomass co-firing. To establish biomass as a viable option, the optimization of the biomass supply chain (BSC) is essential. Although most of the research conducted has focused on optimization models, the purpose of this paper is to incorporate machine-learning (ML) algorithms into a stochastic Mixed-Integer Linear Programming (MILP) model to select potential storage depot locations and improve the solution in two ways: by decreasing the total cost of the BSC and the computational burden. We consider the level of moisture and level of ash in the biomass from each parcel location, the average expected biomass yield, and the distance from each parcel to the closest power plant. The training labels (whether a potential depot location is beneficial or not) are obtained through the stochastic MILP model. Multiple ML algorithms are applied to a case study in the northeast area of the United States: Logistic Regression (LR), Decision Tree (DT), Random Forest (RF), and Multi-Layer Perceptron (MLP) Neural Network. After applying the hybrid methodology combining ML and optimization, it is found that the MLP outperforms the other algorithms in terms of selecting potential depots that decrease the total cost of the BSC and the computational burden of the stochastic MILP model. The LR and the DT also perform well in terms of decreasing total cost.

scholarly journals Learning to Solve Large-Scale Security-Constrained Unit Commitment Problems

2020 ◽  
Author(s):  
Álinson S. Xavier ◽  
Feng Qiu ◽  
Shabbir Ahmed
Keyword(s):  
Machine Learning ◽  
Linear Programming ◽  
Integer Linear Programming ◽  
Mixed Integer Linear Programming ◽  
Large Scale ◽  
Unit Commitment ◽  
Energy Sector ◽  
Mixed Integer ◽  
Data Set ◽  
Computational Performance

Security-constrained unit commitment (SCUC) is a fundamental problem in power systems and electricity markets. In practical settings, SCUC is repeatedly solved via mixed-integer linear programming (MIP), sometimes multiple times per day, with only minor changes in input data. In this work, we propose a number of machine learning techniques to effectively extract information from previously solved instances in order to significantly improve the computational performance of MIP solvers when solving similar instances in the future. Based on statistical data, we predict redundant constraints in the formulation, good initial feasible solutions, and affine subspaces where the optimal solution is likely to lie, leading to a significant reduction in problem size. Computational results on a diverse set of realistic and large-scale instances show that using the proposed techniques, SCUC can be solved on average 4.3 times faster with optimality guarantees and 10.2 times faster without optimality guarantees, with no observed reduction in solution quality. Out-of-distribution experiments provide evidence that the method is somewhat robust against data-set shift. Summary of Contribution. The paper describes a novel computational method, based on a combination of mixed-integer linear programming (MILP) and machine learning (ML), to solve a challenging and fundamental optimization problem in the energy sector. The method advances the state-of-the-art, not only for this particular problem, but also, more generally, in solving discrete optimization problems via ML. We expect that the techniques presented can be readily used by practitioners in the energy sector and adapted, by researchers in other fields, to other challenging operations research problems that are solved routinely.

scholarly journals A Study of Learning Search Approximation in Mixed Integer Branch and Bound: Node Selection in SCIP

AI ◽  
2021 ◽  
Vol 2 (2) ◽  
pp. 150-178
Author(s):  
Kaan Yilmaz ◽  
Neil Yorke-Smith
Keyword(s):  
Machine Learning ◽  
Branch And Bound ◽  
State Of The Art ◽  
Heuristic Method ◽  
Imitation Learning ◽  
Mixed Integer ◽  
Node Selection ◽  
Selection Policy ◽  
State Of Practice ◽  
Optimality Gap

In line with the growing trend of using machine learning to help solve combinatorial optimisation problems, one promising idea is to improve node selection within a mixed integer programming (MIP) branch-and-bound tree by using a learned policy. Previous work using imitation learning indicates the feasibility of acquiring a node selection policy, by learning an adaptive node searching order. In contrast, our imitation learning policy is focused solely on learning which of a node’s children to select. We present an offline method to learn such a policy in two settings: one that comprises a heuristic by committing to pruning of nodes; one that is exact and backtracks from a leaf to guarantee finding the optimal integer solution. The former setting corresponds to a child selector during plunging, while the latter is akin to a diving heuristic. We apply the policy within the popular open-source solver SCIP, in both heuristic and exact settings. Empirical results on five MIP datasets indicate that our node selection policy leads to solutions significantly more quickly than the state-of-the-art precedent in the literature. While we do not beat the highly-optimised SCIP state-of-practice baseline node selector in terms of solving time on exact solutions, our heuristic policies have a consistently better optimality gap than all baselines, if the accuracy of the predictive model is sufficient. Further, the results also indicate that, when a time limit is applied, our heuristic method finds better solutions than all baselines in the majority of problems tested. We explain the results by showing that the learned policies have imitated the SCIP baseline, but without the latter’s early plunge abort. Our recommendation is that, despite the clear improvements over the literature, this kind of MIP child selector is better seen in a broader approach to using learning in MIP branch-and-bound tree decisions.

scholarly journals 102 The deep phenotype characterization of ‘Off-the-Shelf’ CD19-chimeric antigen receptor (CAR) T cells allows to identify their subset complexity and to optimize their manufacturing

2021 ◽  
Vol 9 (Suppl 3) ◽  
pp. A111-A112
Author(s):  
Cristina Maccalli ◽  
Asma Al-Sulaiti ◽  
Mohammed El-Anbari ◽  
Moza Al Khulaifi ◽  
Mohammed Toufiq ◽  
...  
Keyword(s):  
Machine Learning ◽  
T Cells ◽  
T Cell ◽  
Chimeric Antigen Receptor ◽  
Antigen Receptor ◽  
T Cell Subsets ◽  
Mixed Integer ◽  
Car T Cells ◽  
Car T ◽  
Phenotype Characterization

BackgroundUmbilical cord blood (UCB) represents a promising source of T cells for the generation of ‘off-the-shelf’ T cells engineered to express a chimeric antigen receptor (CAR). This study is aimed at understanding the composition of T cell subsets within UCB-CAR-T cells.MethodsT cells, either from UCB or peripheral mononuclear cells (PBMCs) of healthy donors, were activated in vitro with CD3/CD28 mAbs either conjugated to magnetic beads (Dynabeads) or to a colloidal polymeric nanomatrix (TransAct; Miltenyi Biotec). T cells were then transduced with lentiviral vectors encoding for CD19-CD28z or CD19-4-1BBz CARs. The deep phenotype analyses of the CD19-CAR-T cells (N=32) was performed through a multidimensional flow cytometry to assess the expression/co-expression of T cell-associated markers (N=29). The NGFR was utilized as probe for the expression of CD19-CAR. To select the pertinent markers characterising the different groups, we applied a machine learning technique called L0-regularized logistic regression,1 2 and implemented in the R packageL0Learn. 5-fold cross-validation (CV) was used to select the optimal values of the tuning parameters. CD19-CAR-T cells have been also characterized for the transcriptomic profile by parallel quantitative PCR using the high throughput BioMark HD platform and for cytokines, perforin and granzyme B release upon the co-culture with CD19 expressing or not target cells.ResultsT lymphocytes UCB showed efficient expression of the CARs (40–70% of positive cells). Different T cell subsets could discriminate the composition of T cells activated with either Beads or TranAct. CD4+NGFR+CD45RA+ or CD8+NGFR+CD45RA+ T cells associated with different combinations of CCR7, CD62L, LAG3, CD57, CD56 could discriminate between cells activated with Beads vs. TranAct (figures 2–3). CD8+NGFR+CD45RO+CD279−CD152+ T cells were also differentially expressed in TranAct vs. Beads. The PCA analyses also highlighted differences in terms of CD19-CAR-T cell subsets (such as CD8+NGFR+CD45RO+CD62L+, CD8+NGFR+CD45RO+CCR7+, CD8+NGFR+CD45RO+CD272+TIM−3+, CD8+NGFR+CD45RO+CD272+TIM−3+, CD8+NGFR+CD45RA+CD272+TIM−3− and CD4+NGFR+CD45RA+CD272−TIM−3+) in PBMCs vs. UCBs (figure 1). In addition, bystander T cells with different phenotype not expressing the CARs were also detected within the populations of T cells with different origins. Similarly, different T subsets were found in relationship with the sources of T cells. These CD19-CAR-T cells were also characterized for the anti-tumor activity and transcriptomic profiling.Abstract 102 Figure 1PCA of CAR-T cells from UCB vs. PBMCsAbstract 102 Figure 2PCA of CAR-T cells from UCB to compare TransAct vs. beadsAbstract 102 Figure 3PCA of CD19-CAR-T cells to compare TransAct vs. Beads irrespective of the source of the T cellsConclusionsThe combination of deep phenotype characterization with novel statistical tools allowed to identify the complexity of subsets in the engineered T cells in relationship with the starting material and the methods for the activation of the lymphocytes. These findings have important implications for the optimization of the manufacturing of CD19-CAR-T cells.ReferencesAntoine Dedieu, Hussein Hazimeh, and Rahul Mazumder. Learningsparse classifiers: Continuous and mixed integer optimization perspectives. Journal of Machine Learning Research 2021.Hussein Hazimeh and Rahul Mazumder. Fast best subset selection: Coordinatedescent and local combinatorial optimization algorithms. Operations Research 2020;68(5):1517–1537.Ethics ApprovalSidra Medicine’s Ethics Board approval, #1812044429

scholarly journals Mixed Integer Linear Programming based machine learning approach identifiesregulatorsof telomerase in yeast

2016 ◽  
Vol 44 (10) ◽  
pp. e93-e93 ◽  
Author(s):  
Alexandra M. Poos ◽  
André Maicher ◽  
Anna K. Dieckmann ◽  
Marcus Oswald ◽  
Roland Eils ◽  
...  
Keyword(s):  
Machine Learning ◽  
Linear Programming ◽  
Integer Linear Programming ◽  
Mixed Integer Linear Programming ◽  
Mixed Integer ◽  
Learning Approach ◽  
Machine Learning Approach

scholarly journals Optimal training and test sets design for machine learning

2019 ◽  
pp. 1534-1545
Author(s):  
BURKAY GENÇ ◽  
HÜSEYİN TUNÇ
Keyword(s):  
Machine Learning ◽  
Random Sampling ◽  
Prediction Accuracy ◽  
Numerical Study ◽  
Mixed Integer ◽  
Machine Learning Method ◽  
Fixed Size ◽  
Training Set ◽  
Histogram Matching ◽  
Test Sets

In this paper, we describe histogram matching, a metric for measuring the distance of two datasets with exactly the same features, and embed it into a mixed integer programming formulation to partition a dataset into fixed size training and test subsets. The partition is done such that the pairwise distances between the dataset and the subsets are minimized with respect to histogram matching. We then conduct a numerical study using a well-known machine learning dataset. We demonstrate that the training set constructed with our approach provides feature distributions almost the same as the whole dataset, whereas training sets constructed via random sampling end up with significant differences. We also show that our method introduces neither positive nor negative bias in prediction accuracy of a decision tree—used as a representative example of a machine learning method.

Sign in / Sign up

Export Citation Format

Share Document