Predicting future sales using genetic algorithm: data mining sales / Siti Hajar Yaacob
This research is all about predicting future sales using Genetic Algorithm. This is because, most supermarkets have a problem related the product which is exceed expired date and the product out of stock. This problem can cause the imbalance in sales. Based on the problem statement, Genetic Algorith...
Saved in:
| 主要作者: | |
|---|---|
| 格式: | Thesis |
| 語言: | English |
| 出版: |
2014
|
| 主題: | |
| 在線閱讀: | https://ir.uitm.edu.my/id/eprint/64852/2/64852.pdf |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| 總結: | This research is all about predicting future sales using Genetic Algorithm. This is because, most supermarkets have a problem related the product which is exceed expired date and the product out of stock. This problem can cause the imbalance in sales. Based on the problem statement, Genetic Algorithm was applied to test whether it is compatible to predict the future sales. The objectives of this research are data mining are used to extract useful data from sales and product. Besides that, Genetic Algorithm is applying to predict future sales based on the sales and product data. The product is evaluated based on decision tree to determine whether the product need to order or vice versa. Tool decision tree calculate the stock based on the number of product order. This research using Simple Genetic Algorithm which is involved crossover and fitness. The fitness function calculates based on the total sales of the quantity product sales by week. The crossover are combines all the week to make other population. The main purpose of this research to test whether Genetic Algorithm can use to predict future sales and be able generates the graph. The graphs were shows about sales in month. The value of the graph is quantity product sold against the name of the product. Based on the result that has made, 50% state that this system hard to use and not user friendly. However, this system analysis help them to reduce the product that exceed expired date and out of stock. The next aim of this research is the system will be able to give a factor about the changes of sales. |
|---|