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.

In this work, a probabilistic bid generation problem with the pricing of a bundle of lanes and carrier’s vehicle routing is considered as it is an importation in transportation service procurement. Depending on the network of the vehicle, there exist multiple lanes for traveling between two locations. To solve the bid generation problem efficiently, a two-phase method approach is presented. At the core of the procedure a feasible vehicle routing problem on a multidigraph is solved by an exhaustive search algorithm to enumerate all routes concerning routing constraints and treat each route as a decision variable in the set partitioning formulation. We examine our model both analytically and empirically using a simulation-based analysis.

Companies are frequently confronted with the need to order different types of items from a single supplier or to manufacture the items in a production line. Indeed, coordinated ordering of multiple items may lead to important savings whenever a family of items can be ordered from a common supplier, produced in a common facility, or use a common mode of transportation. The Joint Replenishment Problem (JRP) tackles the coordinated replenishment of multiple items by minimizing the total cost, composed of ordering (or setup) costs and holding costs, while satisfying the demand. On the other hand, when items are subject to obsolescence, they may face an abrupt decline in demand as they are no longer needed. This decline can be caused by reasons such as rapid advancements in technology, going out of fashion, or ceasing to be economically viable. The present article develops an extension of the JRP where the items may suddenly become obsolete during an infinite planning horizon. The point at which an item becomes obsolete is uncertain. The lifetimes of the items are assumed to follow independent negative exponential distributions. A model is proposed by using the total expected discounted cost as the minimization criterion. The time value of money is considered through an appropriate discount rate. Extensive tests were performed to assess the impact of obsolescence rates and discount rates on the ordering policies. The progressive increase of the obsolescence rates determines smaller periods between successive replenishments, while the progressive increase of the discount rate determines smaller lot sizes.

This paper considers an unrelated parallel machine scheduling problem with variable maintenance based on machine reliability to minimize the maximum completion time. To obtain the optimal solution of small-scale problems, we firstly establish a mixed integer programming model. To solve the medium and large-scale problems efficiently and effectively, we develop a hybrid discrete spider monkey optimization algorithm (HDSMO), which combines discrete spider monkey optimization (DSMO) with genetic algorithm (GA). A few additional features are embedded in the HDSMO: a three-phase constructive heuristic is proposed to generate better initial solution, and an individual updating method considering the inertia weight is used to balance the exploration and exploitation capabilities. Moreover, a problem-oriented neighborhood search method is designed to improve the search efficiency. Experiments are conducted on a set of randomly generated instances. The performance of the proposed HDSMO algorithm is investigated and compared with that of other existing algorithms. The detailed results show that the proposed HDSMO algorithm can obtain significantly better solutions than the DSMO and GA algorithms.

Operations with Electric Vehicles (EVs) on logistic companies and power utilities are increasingly related due to the charging stations representing the point of standard coupling between transportation and power networks. From this perspective, the Multi-depot Electric Vehicle Routing Problem (MDEVRP) is addressed in this research, considering a novel hybrid matheheuristic approach combining exact approaches and a Chu-Beasley Genetic Algorithm. An existing conflict is shown in three objectives handled through the experimentations: routing cost, cost of charging stations, and increased cost due to energy losses. EVs driving range is chosen as the parameter to perform the sensitivity analysis of the proposed MDEVRP. A 25-customer transportation network conforms to a newly designed test instance for methodology validation, spatially combined with a 33 nodes power distribution system.

Reverse auction is an effective tool for a 4PL to purchase transportation services. This paper investigated a new transportation services procurement problem for 4PL, which involves three features: the 4PL’s loss-averse behavior, price and non-price attributes, and multiple transportation requests. An efficient multi-attribute multi-item reverse auction mechanism considering the 4PL ex-ante and ex-post satisfaction (EES-MMRA) is proposed to purchase transportation services for the 4PL. In the EES-MMRA, integrating the allocation rule with the 4PL ex-ante satisfaction, a 0-1 programming model is constructed to determine winning 3PLs and obtain efficient allocations. Then, a payment rule considering the 4PL ex-post satisfaction is established to ensure truthful bidding of 3PLs. And we discuss some desirable properties (e.g., incentive compatibility, individual rationality, efficiency, and budget balance properties) to justify the EES-MMRA mechanism, subsequently. Next, several numerical experiments are conducted to demonstrate the effectiveness and applicability of the EES-MMRA mechanism. Furthermore, sensitivity analysis presents the influences of the weights of the non-price attributes, risk attitude coefficients, and loss aversion coefficients. Finally, we conduct comparison analysis to show the advantages of the EES-MMRA mechanism over the known Vickrey–Clark–Groves (P-VCG) mechanism.

Training and productivity have been found positively correlated in theory and practice. The pattern of the relationship, however, was not explored in pre - and post-regulation in the conceived sector. This paper aims to observe the effects of regulation and training expenses along with other covariates on employee productivity in the Microfinance sector of Bangladesh. Using annual data of MRA-licensed MFIs, we have estimated both panel and cross-sectional regression models. The regression results confirm the theoretical relationship between training expenses and employee productivity. The regulation also worked positively in enhancing employee productivity. However, in the early stage of regulation, the average productivity gain due to regulation was substantial and was showing an increasing trend but then it declined and reached a constant level of about 4% - 5% each year. Between 2008 and 2011, both regulation and training positively contributed to the gains in average productivity of the employees. After 2012, there was a positive trend of average productivity elasticity of training expenses but there was a flat effect of regulation after 2012. Regulation was found to short-run shifter in the average productivity of employees while training expenses had a positive trend effect.

The main focus of this research paper is to shed light on the challenges encountered by the chair production department employees when it comes to identifying and locating glue residue after the assembly process. This issue arises due to the transparent nature of the glue, making it difficult to detect and subsequently negatively impacting the overall quality of the chairs in the polishing department. To tackle this problem, various experiments and tests were conducted, which are thoroughly discussed in this paper. The findings of this study are not only applicable to the specific manufacturing company under investigation but also to other companies within the industry that face similar difficulties in detecting defects caused by transparent glue. As a result, a new method for detecting glue defects is proposed, which can be adopted by various industries encountering similar challenges.

Effective application of tools and techniques is critical for successful Total Quality Management (TQM) implementation. The influence of organization profile on the application of TQM tools and techniques and outcomes is investigated in Indian organizations. Correlation analysis and ANalysis Of VAriance (ANOVA) are utilized for the research. Tools and techniques are found having significant relation to the outcomes of implementation. Organization’s age, size, and turnover significantly influenced the TQM results. Organization’s size and turnover significantly influenced the application of tools and techniques. The study suggests taking the organization profile into consideration for effective implementation. Further research can be conducted in other countries.

The COVID-19 surge has mostly affected people and wreaked havoc on multiple sectors of the global economy. This study uses artificial neural networks (ANN) to develop COVID-19 prediction models to minimize the perilous situation. With positive infection data, these hybrid artificial neural network models looked at COVID-19 cases in Andhra Pradesh, India. Then, COVID-19 data were divided into training and testing for simulation. The developed model that takes the previous 14 days into account outperforms the others, depending on the results. According to the developed ANN models for Andhra Pradesh districts, the prediction model that works well and yields positive results is the one that receives lower values of error metrics like MSE, RMSE, MAE, and MAPE and higher values of R2. The hybrid neural network model that considers the previous 14 days for prophecy has suggested anticipating daily positive suffering, notably in areas of Andhra Pradesh, as a result of the collected data. Linear regression, ARIMA, and LSTM have been used for model assessment. The proposed 14-day model statistically surpasses all metrics in RMSE, MAE, MAPE, and R2. This study showed that an ANN-based model can predict the COVID-19 outbreak as well as other epidemics.