There have been many applications for machine learning algorithms in different fields. The importance of hyperparameters for machine learning algorithms is their control over the behaviors of training algorithms and their crucial impact on the performance of machine learning models. Tuning hyperparameters crucially affects the performance of machine learning algorithms, and future advances in this area mainly depend on well-tuned hyperparameters. Nevertheless, the high computational cost involved in evaluating the algorithms in large datasets or complicated models is a significant limitation that causes inefficiency of the tuning process. Besides, increased online applications of machine learning approaches have led to the requirement of producing good answers in less time. The present study first presents a novel classification of hyperparameter types based on their types to create high-quality solutions quickly. Then, based on this classification and using the hypergradient technique, some hyperparameters of deep learning algorithms are adjusted during the training process to decrease the search space and discover the optimal values of the hyperparameters. This method just needs only the parameters of the previous two steps and the gradient of the previous step. Finally, the proposed method is combined with other techniques in hyperparameter optimization, and the results are reviewed in two case studies. As confirmed by experimental results, the performance of the algorithms with the proposed method have been increased 36.62% and 23.16% (based on the best average accuracy) for Cifar10 and Cifar100 dataset respectively in early stages while the final produced answers with this method are equal to or better than the algorithms without it. Therefore, this method can be combined with hyperparameter optimization algorithms in order to improve their performance and make them more appropriate for online use by just using the parameters of the previous two steps and the gradient of the previous step.

An inventory control system having multiple items in stock is developed in this paper to optimize total cost of inventory and space requirement. Inventory modeling for both the raw material storage and work in process (WIP) is designed considering independent demand rate of items and no volume discount. To make the model environmentally aware, the equivalent carbon emission cost is also incorporated as a cost function in the formulation. The purpose of this study is to minimize the cost of inventories and minimize the storage space needed. The inventory models are shown here as a multi-objective programming problem with a few nonlinear constraints which has been solved by proposing a meta-heuristic algorithm called multi-objective particle swarm optimization (MOPSO). A further meta-heuristic algorithm called multi-objective bat algorithm (MOBA) is used to determine the efficacy of the result obtained from MOPSO. Taguchi method is followed to tune necessary response variables and compare both algorithm's output. At the end, several test problems are generated to evaluate the performances of both algorithms in terms of six performance metrics and analyze them statistically and graphically.

There is an acute need to evaluate the energy consumption of buildings in response to climate change. The “occupant” factor has been largely overlooked in building energy analysis. This research aims at investigating occupancy existence in the office environment using a hybrid artificial neural network with metaheuristic algorithms for improved energy management. It proposes and compares three classification models, namely particle swarm optimization (PSO), gravitational search algorithm (GSA), and hybrid PSO-GSA in combination with the feedforward neural network (FFNN). The inputs to these models are data related to temperature, humidity, light, and carbon dioxide emissions. Two data sets are used for testing the models while the office door is open and closed. The capabilities of the optimized models are evaluated using best, average, median, and standard deviation of the mean squared error. Most of the performance metrics indicate that the FFNN-PSO-GSA model exhibits better performance compared to the other models using the two datasets. The proposed model yields a classification accuracy ranging between 98.47-98.73% using one predictor (i.e., temperature). Besides, it yields an accuracy ranging between 85.45-94.03% using temperature and CO2 predictors. It can be concluded that the FFNN combined with PSO and GSA algorithms can be a useful tool for occupancy detection modeling.

Inventory control of deteriorating items constitutes a large part of the world’s economy and covers various goods including any commodity, which loses its worth over time because of deterioration and/or obsolescence. Vendor managed inventory (VMI), which is a win-win strategy for both suppliers and buyers gains better results than traditional supply chain. In this research, we study an economic order quantity (EOQ) with shortage in form of partial backorder under VMI policy. The model is concerned with multi-item subject to multi-constraint including storage space, time period and budget constraints. Two metaheuristic algorithms, namely Simulated Annealing and Tabu Search, are used to find a near optimal solution for the proposed fuzzy nonlinear integer-programming problem with the objective of minimizing the total cost of the supply chain. Furthermore, the sensitivity analysis of the metaheuristic parameters is performed and five numerical examples containing different numbers of items are conducted in order to evaluate the performance of the algorithms.

Locating, routing and inventory control in production and distribution centers are the most important decisions in supply chain management. Because of the dependence of these decisions to each other, considering these three aspects simultaneously can have a huge impact on cost reduction. In this study, first, a location-inventory model is developed by utilizing METRIC approach and then, METRIC approach is applied to the location-inventory-routing model. The intended supply chain includes a supplier, distributors and retailers, and the inventory control policy is implemented for both the distributors and retailers. Retailers' demand follows Poisson distribution and the lead-time is also considered probabilistic and is affected by the shortage in distribution centers. Given that the presented model belongs to the class of NP-hard problems, a hybrid metaheuristic solution method is also presented to solve the resulted problem. The proposed hybrid metaheuristic algorithm contains a Simulated Annealing algorithm, to optimize the location-routing problem, and a Genetic Algorithm, to optimize the inventory problem. Also, to evaluate the performance of hybrid algorithm, a comparison between the results of the proposed hybrid algorithm and the exact solutions obtained from Lingo software is provided and, finally, the results are analyzed.

Multi-objective optimization is an optimization problem with some conflicting objectives to be attained, simultanously. This paper reviewed literature about multi-objective optimization problems for supply chain management. The review aimed to provide the lastest research views and recomendations for further studies. We discussed the lastest ten years publications about multi-objective optimization for supply chain management. The scope of this review was classified into five categories i.e. problem statements, multi-objective frameworks, mathematical formulation modeling, optimization techniques, and representation of supply chain. Multi-objective optimization approaches, both classical and metaheuristic approaches, were discussed, accordingly. In this review, we conducted conclusion and recomendations about likelihood research directions in future.