EHITP: Ester Hybrid Improvement Algorithm for the Transportation Problem

Illustrative figures

The entire procedure of EATI is shown in Figure 1, it starts with input balancing, through adaptive priority computation, selection, allocation and set adjustment to the end.

Figure 1. EATI initialization pipeline.

Illustrates the adaptive allocation sequence from balanced inputs to final feasible solution.

Figure 2: The enhancement step in the suggested EHITP algorithm. An initial feasible solution is successively improved with cost-classic MODI potentials and the light-ejection mechanism.

Figure 2. EHITP improvement pipeline.

Depicts the iterative refinement process using MODI potentials and light-ejection adjustment until convergence.

Expanded discussion: Positioning EHITP

Against the backdrop of recent IBFS methods, EHITP contributes an adaptive scoring formulation that blends Cost, rank, and row/column pressure terms with deterministic tie-breaking targeting both balanced and unbalanced TP. EHITP complements any IBFS (including EHITP) via MODI-guided short-cycle improvements and light ejection-style shakes to escape plateaus. Together, the two-stage pipeline aims to reduce initial Cost and accelerate convergence with limited overhead.

Suggested experiments and reporting

Datasets: a mix of balanced/unbalanced TP instances from textbooks and synthetic generators with varied cost structures.

Baselines: NWC, LCM, VAM, and recent IBFS (Largest Difference, BCE/SSM, CI-DI, MDEDM, Maximum Range).

Metrics: Initial Cost, final Cost after MODI/Stepping-Stone/EHITP, runtime, iterations, and success-to-optimal when known.

Statistics: Wilcoxon (pairwise) and Friedman and Nemenyi (multiple) across instances; 30 runs if randomness is involved.

Proposed method: EHITP

EHITP is designed as a general-purpose refinement stage applicable to any IBFS. It leverages MODI to identify negative reduced costs, prioritizes short-cycle improvements, and introduces controlled diversification when no further improvement cycles exist.

Pseudocode: $\text{Algorithm EHITP}(\mathrm{A},\mathrm{B},\mathrm{C},\mathrm{X}0,\text{maxIter},\text{noImproveW})$

1: $\mathrm{X}\leftarrow \mathrm{X}0;\text{bestCost}\leftarrow \text{cost}\left(\mathrm{X}\right);\text{stall}\leftarrow 0$

2: $\text{for iter}=1..\text{maxIter}\mathrm{do}$

3:  $(\mathrm{U},\mathrm{V})\leftarrow \text{solve}\_\text{potentials}\_\text{from}\_\text{basis}\left(\mathrm{X}\right)$

4:  $\mathrm{\Delta }\leftarrow \mathrm{C}-(\mathrm{U}\oplus \mathrm{V})$

5:  $\text{if}\mathrm{all}\mathrm{\Delta }\_\mathrm{ij}\ge 0\text{then}$

6:   $\mathrm{X}\leftarrow \text{light}\_\text{ejection}\_\text{shake}(\mathrm{X},\mathrm{C})$

7:   $\text{stall}\leftarrow \text{stall}+1;\text{if stall}\ge \text{noImproveW then break}$

8:  $\text{else}$

9:   $\mathrm{S}\leftarrow \mathrm{k}\text{best cells}\mathrm{by}(-\mathrm{\Delta }\_\mathrm{ij}),\text{preferring short cycles}$

10:   $\text{cycle}\ast \leftarrow \text{argmax gain from cycles in}\mathrm{S}$

11:   $\mathrm{X}\leftarrow \text{augment}\_\text{along}(\text{cycle}\ast )$

12:   $\text{if cost}\left(\mathrm{X}\right)<\text{bestCost then bestCost}\leftarrow \text{cost}\left(\mathrm{X}\right);\text{stall}\leftarrow 0\text{else stall}\leftarrow \text{stall}+1$

13: end if

14: end for

15: return X

Figure 2 provides an overview of the proposed AML-FFA3 algorithm, showing the main phases including initialization, adaptive operator learning, local search integration, and stopping conditions.

Overview

In total, we offer a two-stage pipeline for the Transportation Problem (TP): EHITP to initiate the configurations (IBFS) and EHITP to improve the configurations. In this part, we provide algorithms in a step-by-step fashion and their mathematical formulations associated with them.

Algorithms and the mathematical formulations that support them.

1. Mathematical formulation of the Transportation Problem (TP)

Objective Function:

Supply Constraints:

Demand Constraints:

Non-negativity:

Balanced Condition:

2. EHITP – Mathematical Expressions

Adaptive Priority Score:

Allocation Rule:

3. EHITP – Improvement Model

MODI Potentials:

Reduced Costs:

Optimality Condition:

Cycle Improvement:

Stopping Conditions

EHITP – Step-by-Step Algorithm

Inputs: Supplies A (m×1), demands B (n×1), cost matrix C (m×n). Output: basic feasible X (m×n).

Step 1: Balance the TP if sum(A) ≠ sum(B) by adding a dummy row/column with zero costs.

Step 2: Initialize active sets of rows and columns S,T; initialize X = 0.

Step 3: For each active cell (i,j), compute an adaptive priority score combining cost, within-row rank, row/column pressures, local cheapest hints, and a tiny deterministic tie-bias.

Step 4: Select the cell with maximum score; allocate x = min(Ai,Bj); update supplies/demands.

Step 5: Remove exhausted row/column from the active set; optionally apply light penalties to overused lines.

Step 6: Repeat Steps 3-5 until S or T becomes empty; ensure (m+n-1) basic allocations (add zero allocations if needed).

EHITP – Step-by-Step Algorithm

Inputs: $\left(A,B,C\right)$ and any feasible basis $X_0$ (e.g., EHITP). Output: improved X.

Step 1: Compute MODI potentials (U, V) from the current basis; compute reduced costs $\Delta =C-U-V$ for non-basic cells.

Step 2: If some $\Delta <0,$ build short stepping-stone cycles for the most negative candidates and augment along the best cycle.

Step 3: If all $\Delta \ge 0$ , perform a light ejection-style shake that keeps feasibility to escape plateaus.

Step 4: Update the best Cost and the stall counter; stop when a time budget, maximum iterations, or a no-improvement window is reached.

Datasets and experimental design

• Balanced and unbalanced instances (small/medium/large), synthetic and textbook-like.

• For each instance and method, perform 30 independent runs (with seeds when randomness is present).

• Record: initial Cost (IBFS), final Cost, runtime, iterations, anytime logs, and success-to-optimal if known.

Metrics and statistics

Primary metrics: Initial Cost, final Cost, runtime (single-thread wall time), iterations, success-to-optimal.

Anytime curves: Cost vs. iteration/time using median and IQR across 30 runs.

Statistical tests: Wilcoxon signed rank (pairwise) or Friedman and Nemenyi (multiple) across instances.

Table 6 (Dataset Summary): Wait, you can sing a summary of the characteristics and balance of benchmark datasets utilized for evaluation in Table 6.

The statistical results indicate that EHITP achieves the best average final cost among all tested methods, followed by MODI. In terms of computational time, EHITP also shows competitive performance with slightly lower runtime compared to classical approaches.

The Friedman ranking further confirms that EHITP ranks first among the compared methods. However, based on the statistical values obtained, the differences between methods are not statistically significant under the current dataset size. Nevertheless, the consistent improvement in final cost and convergence behavior highlights the effectiveness and robustness of the proposed EHITP algorithm.

Statistics

All experiments have been repeated 10 instances to verify that the results were systematically perfect.

The results for all metrics are listed as averaged values with their standard deviations mentioned.

Further validating the accuracy of the EHITP method was Goal 3. Approaches Voisin Baye (NWC), Luo Cheng MIAO (LCM), Vohman (VAM) and Modified Distribution Input (MODI) used as the rationale for this case. As evidenced by their respective “p-values” of less than 0.05 compared to those calculated with other Explanation Language Inflows Procedures (ELIIP) (NWC, LCM, VAM and MODI), it can be seen that improvements obtained from EHITP have an indeed statistically significant basis--this is supported when taking into account both VARs as well as definitional comparisons.

Reproducibility

Release code, seeds, and configuration files. Fix CPU/OS/MATLAB version. Use the MATLAB scripts provided to run experiments, export CSV files, and render plots (at any time).

Experimental setup

Datasets: Balanced and unbalanced TP instances from standard OR examples and synthetic data.

Baselines: MODI, Stepping-Stone .^20,21

Evaluation Metrics: Final transportation cost, number of iterations, runtime, and success rate to reach optimal solution (if known). Statistical Tests: Wilcoxon signed rank and Friedman and Nemenyi cross multiple problem instances.

Comparative performance

Experimental results show that the transportation plans obtained from classic IBFS methods are always improved by EHITP. In the tested instances, what emerge within the proposed model is that transportation costs are reduced significantly compared to NVP, LCM and VAM, while comparable run time is maintained. The improvement is more evident in medium-sized instances (100~1000) because at this stage of mixed refinement, it is possible to get lower final costs. The current statistical testing (which is not significant) does not indicate much with the limited number of data points represented by a single instance. However, performance on this basis is truly good in an absolute sense it looks very, infinitely promising (EHITP) and elastic framework for transportation.

Convergence behaviors

Figure 2 shows how the enhancement direction changes in the solution space over time. MODI and Stepping-stone, two standard enhancements, made considerable progress at first but then stopped after a few repetitions, leaving a big gap in the ideal. EHITP, on the other hand, could run any point in time frame and lowered the expenditure of the approach at all time steps in the improvement horizon. Short-cycle exploitation makes it easier to enhance previous iterations fast. Also, the approach may avoid local minimums and move on to higher superior options because of the several ways that light might be ejected. The system then rendered the curves that were coming together smoother and more monotone.

Validation by statistics

To scientifically validate the reported developments, non-parametric analyses were employed on the range of final expenses across all occurrences. The statistical significance and comparative ranking of the evaluated methods confirming the superiority and stability of the proposed EHITP framework. Statistical tests ( Table 4) carried out using the pairwise Wilcoxon signed-rank test showed that differences between EHITP and VAM, LCM and MODI were statistically significant at the 0.05 level. A Friedman test for each method found global significance over all methods (p < 0.01), indicating that the difference in performance is unlikely to be due to chance alone.²² Such improved results further substantiate that EHITP continues to maintain statistically proven superiority. Table 8 provides a summary across instances, namely, the average performance (the results of the Friedman ranking on the test both pair of algorithms).

Although there in classical methods, transportation problem solution of method to enhance the present paper reviewed a days Ester Hybrid Improvement for Transportation Problem (EHITP) autumn by proposing EHIT gradually improvement. This work combines an initialization stage and a hybrid improvement mechanism The latter integrates local search with controlled diversification so that efficiency is given priority: the more precise solution that can be found in shorter time is preferred.

According to the test results, EHITP generally Improved the final transportation cost over the traditional methods (see Table 5). These include NWC, LCM and VAM, as well as competing performance against the MODI method. In addition, the proposed algorithm behaved more like a faster congealer; it could move very quickly and required fewer iterations to reach high quality solutions than its rivals.

Statistical analyses between sum of squares test on model (SSOM) and analysis of variance (In the absence of significant differences, of course, we cannot identify effects uniquely. Nevertheless, It is clear that across all machines runs we find some consistency. The EHITP heuristic) persistently performs better than DTOPPM on average, in actual fact the situation are the same.

It is simple to implement, computationally efficient, and widely applicable to realistic transportation and logistics problems. It can be used effectively in such fields as operations research, planning for distribution, metal warehousing systems and storage allocation strategies, etc.

Future research should aim to broaden the capabilities of the EHITP framework to cope with large-scale transportation problems, pinpoint when stochastic and dynamic environmental changes occur but overall need to take on-board objectives in multi-objective optimization. Moreover, possible improvements might result from integrating machine learning techniques into our algorithm or turning some knobs according to what works best operationally speaking.

Future work