Search

This paper presents a novel variant of the teaching–learning-based optimization algorithm, termed BLTLBO, which draws inspiration from the blended learning model, specifically designed to tackle high-dimensional multimodal complex optimization problems. Firstly, the perturbation conditions in the "teaching" and "learning" stages of the original TLBO algorithm are interpreted geometrically, based on which the search capability of the TLBO is enhanced by adjusting the range of values of random numbers. Second, a strategic restructuring has been ingeniously implemented, dividing the algorithm into three distinct phases: pre-course self-study, classroom blended learning, and post-course consolidation; this structural reorganization and the random crossover strategy in the self-learning phase effectively enhance the global optimization capability of TLBO. To evaluate its performance, the BLTLBO algorithm was tested alongside seven distinguished variants of the TLBO algorithm on thirteen multimodal functions from the CEC2014 suite. Furthermore, two excellent high-dimensional optimization algorithms were added to the comparison algorithm and tested in high-dimensional mode on five scalable multimodal functions from the CEC2008 suite. The empirical results illustrate the BLTLBO algorithm's superior efficacy in handling high-dimensional multimodal challenges. Finally, a high-dimensional portfolio optimization problem was successfully addressed using the BLTLBO algorithm, thereby validating the practicality and effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

Abstract. Buyer finance has been practiced by manufacturers/assemblers for years; however, few papers have investigated the efficacy of buyer finance in an assembly system with multiple suppliers. This paper fills [...]

This paper presents an analytical model for sellers with finite capacity to optimise their price and maximum deal size limit on group-buying websites. For the general demand functions that satisfy some mild regularity conditions, the optimal strategies and the corresponding deal parameters are characterised. The optimal strategies demonstrate that deep online discounts (selling products below the cost or even selling products for free) can be optimal if the maximum deal size is used strategically. Moreover, deep discounts can be beneficial even if the capacity is tight, which explains the use of deep discounts commonly observed in practice. Our results also provide the operators of group-buying websites with useful suggestions on how to induce sellers to offer deep discounts. Sensitivity analysis with regard to the minimum deal size and capacity is provided. Our analysis shows that counter-intuitively selling out capacity may not always be optimal, even if the amount of capacity cannot satisfy the unconstrained optimal sales quantity online. Finally, we extend the model to consider offline prices being sellers' decisions and discuss the robustness of the optimal strategies when the demand is stochastic. [ABSTRACT FROM AUTHOR]

The credit limit is the maximum amount of accounts payable (AP) that a firm owes to its supplier. This limit establishes a tie between the effective purchasing cost and the order quantity as well as the AP level. We present a multi‐period stochastic inventory model with payment delay and credit limit. Our model aims to capture the property of the optimal ordering policy with respect to the AP and inventory states. For the backlogging and lost‐sales cases, we present the method to reduce the dimension of state space. In both cases, the optimal order quantity is demonstrated to be decreasing in the AP and inventory level of each age. In addition, the sensitivity of the optimal order quantity with respect to the AP decreases with age. That is, the optimal order quantity is more sensitive to the younger AP. In the backlogging case, the base‐stock policy is not optimal due to the presence of credit limits. Somewhat surprisingly, our numerical example shows that the optimal order quantity is not necessarily increasing in the credit limit. Finally, based on the L♮‐convexity of the cost‐to‐go function, we design a linear programming greedy heuristic policy for the infinite‐horizon version of our model with independent and identically distributed demands. Numerical results show that our heuristic policy performs well. [ABSTRACT FROM AUTHOR]

The convention of selling on credit (to customers) results in mass accumulation of accounts receivable (AR) on the balance sheet of firms. However, capital‐constrained firms do not have enough capital to invest in AR and cover the production cost incurred during the credit period. To finance future business, a capital‐constrained firm can employ factoring—a financing scheme wherein firms sell AR to a financial institution (called a factor) at a discount—to advance cash from AR. By formulating a time‐continuous model with constant demand over an infinite horizon, we study the factoring policy of firms in two practice‐based discounting schemes: automatic discounting and manual discounting. In the automatic discounting scheme, AR should be discounted at the same age, whereas this requirement is relaxed in the manual discounting scheme and how long in advance to discount the AR is contingent over time. In both discounting schemes, the firm needs to choose the timing of discounting in order to reach a capital‐unconstrained state as soon as possible. In the automatic discounting scheme, we approximate the firm's objective function with a quasi‐convex function whose error is demonstrated to be small. Based on this approximation, the firm's optimal decision and the factor's profit can be calculated more easily. When manual discounting is adopted, the firm should meet all the demand by exploiting factoring if the profit margin under immediate discounting is nonnegative. Given the same factoring discount rate, manual discounting is always more attractive to the firm than automatic discounting. However, the preference of the factor over the two discounting schemes depends on the factoring discount rate. [ABSTRACT FROM AUTHOR]