代写IOM103 Artificial Intelligence in Business代写Java程序
- 首页 >> CSCoursework: Optimizing Inventory Management with 1D Bin Packing
Artificial Intelligence in Business (IOM103)
1 Overview
This coursework focuses on the application of the 1D bin packing problem to inventory management in a retail or manufacturing setting. Students will be tasked with developing an algorithm to efficiently allocate items of various sizes into fixed-size bins (representing storage units or shipping containers) to minimize space wastage and optimize inventory storage or distribution costs. This real-world applica- tion helps in understanding how computational algorithms can solve operational challenges in inventory management.
2 Background
In inventory management, space optimization is crucial for reducing storage and transportation costs. The 1D bin packing problem, a classic problem in artificial intelligence and management, involves packing objects of different volumes into a finite number of bins with a certain capacity in the most space-efficient way. Applying this concept to inventory management can significantly enhance operational efficiency and cost-effectiveness.
3 Bin Packing Problem (BPP)
Given a set of n items, each item j has a volume of aj , BPP aims to pack all items in the minimum number of identical-sized bins without violating the capacity of bins (V). The problem can be mathematically formulated as follows:
This mathematical formulation is generally NOT solvable by existing integer programming solvers like CPlex, Gurobi, and LPSolve, especially when the number of items n is large. To consistently solve the problem with good quality solutions, metaheuristics and hyper-heuristics are used, which is the task of this coursework.
4 Objectives
. To understand the principles and applications of the 1D bin packing problem in inventory manage- ment.
. To develop and implement an algorithm that optimizes the packing of items into bins.
. To evaluate the algorithm’s efficiency and effectiveness in reducing unused space and potential cost savings.
5 Materials
Students will be provided materials that include:
. A list of items to be stored or shipped, each with its volume.
. The volume capacity of the bins (assumed to be uniform for simplicity)
. An example python code can output the solution file.
JSON File: https://core.xjtlu.edu.cn/pluginfile.php/178458/mod_folder/content/0/CW_ ins.json?forcedownload=1
Example Code: https://core.xjtlu.edu.cn/pluginfile.php/178458/mod_folder/content/0/CW_exp.py?forcedownload=1
6 Tasks
. Literature Review:
Study the 1D bin packing problem and existing algorithms used to solve it, such as First-Fit, Best-Fit, and First-Fit Decreasing algorithms.
. Algorithm Selection and Development:
Choose an appropriate algorithm or develop a new strategy for solving the 1D bin packing problem in the context of inventory management. Implement the algorithm using a programming language of choice (e.g., Python, Java).
. Simulation and Optimization:
Run simulations using the provided dataset to allocate items into bins based on the developed algorithm. Optimize the algorithm to improve the space utilization ratio.
. Analysis and Reporting:
Analyze the algorithm’s performance, comparing it with other strategies studied during the literature review. Prepare a detailed report documenting the project’s methodology, implementation, results, and conclusions.
7 Deliverables
. A no more 3 pages report includes a declaration of AI usage, introduction, description of the chosen algorithm, implementation details, results, analysis (compare with other algorithms), and conclusions (a PDF file).
. Source code for the implemented algorithm (a Python file program, program should take no more than 5 mins to produce output).
. Solutions for the provided problem sets (a JSON file output by example code).
Note:
. Submit your lab report via the dedicated Learning Mall coursework link (https://core.xjtlu . edu.cn/mod/coursework/view.php?id=127472).
. Essay only accepts PDF format.
. Code only accepts Python format.
. Solution only accepts JSON format.
. Please name your submission file as ID FirstName LastName.xxx (e.g., 1234567 Xinan Chen.pdf) .
. Late submission is subject to normal late submission penalties. Three extra hours are provided on the learning mall submission portal in case of internet connection issues. The extra hours should not be relied on for submission. Submitting your essay a few hours before the deadline would always be good practice.
. You can ask your classmates for help, but you cannot completely copy their code, solutions, or essays.
. The PCs in ES209 are used to test your program. If your runtime exceeds 5 minutes, it may result in penalties, or you may even receive a 0 for the results part.
. Your solution should match your program. If your program does not match your solution, you may
receive a 0 mark in the results part.
8 Mark Criteria
1. The quality of the experimental results (40%). Your algorithm shall be tested for a file containing 10 instances chosen from the provided set of instances. The performance of your algorithm is evaluated by computing the absolute gap with the best-known results using:
Best-known results for first 3 instances:
. Instance: instance 1 Best known of bins used: 373
. Instance: instance 2 Best known of bins used: 966
. Instance: instance 3 Best known of bins used: 652
2. The quality of codes (20%) The code quality evaluation will be based on the following criteria, each contributing to the total mark of 20. The objective is to assess not only the functionality of the code but also its efficiency, readability, and adherence to good programming practices.
(a) Readability (10 Marks)
. Clarity: Code is easy to read and understand (5 Marks).
. Comments: Adequate comments are provided to explain the logic and functionality (5 Marks).
(b) Structure and Organization (5 Marks)
. Modularity: Code is well-structured and divided into functions or modules where appro- priate (3 Marks).
. Consistency: Consistent naming conventions and code layout (2 Marks).
(c) Efficiency and Performance (5 Marks)
. Algorithm Efficiency: The chosen algorithm or approach efficiently solves the problem (3 Marks).
. Resource Management: Efficient use of computational resources, minimizing unnecessary computations (2 Marks).
3. Report (40%) The report submitted for evaluation will be assessed based on the following criteria, totaling 40 marks:
(a) Introduction and Background (8 Marks)
. Clarity of Problem Statement (4 Marks): Clear presentation of the problem being ad- dressed.
. Context and Relevance (4 Marks): Explanation of the importance and applications of the problem.
(b) Methodology (8 Marks)
. Approach and Justification (4 Marks): Detailed description of the approach taken to solve the problem, including justification for the chosen methods.
. Implementation Details (4 Marks): Comprehensive explanation of how the solution was
implemented, including any algorithms, models, or frameworks used.
(c) Results and Analysis (16 Marks)
. Presentation of Results (4 Marks): Clear and accurate presentation of the findings, sup- ported by appropriate figures, tables, or graphs.
. Critical Analysis (12 Marks): In-depth analysis of the results, including comparing their
performance with other methods and the pros and cons of each method.
(d) Conclusion and Recommendations (8 Marks)
. Summary of Key Findings (4 Marks): Concise summary of the main outcomes of the project.
. Future Work and Recommendations (4 Marks): Insightful suggestions for future research or improvements to the project based on the findings.
9 AI Usage
This coursework permits using AI to assist in programming or essay writing. However, you must declare the use of AI at the beginning of your essay (e.g., used AI for programming). Failure to do so may result in a loss of marks for any part where AI assistance is detected.
10 Repetition Check
If a repetition rate higher than 20% is found (or 5% from a single source), your submission will be checked by lecturers. You may lose a certain number of marks for improper citations, plagiarism, etc. Severe cases will be reported to the exam officer for further action.
11 Submission Deadline
31st May 2024, 4pm Beijing Time