Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Steel product spot pricing system and method based on machine learning

A machine learning and product technology, applied in the computer field, can solve problems such as inability to accurately and timely grasp market dynamics, market competition and emergencies are not flexible and timely, and limit the rationalization of steel sales.

Pending Publication Date: 2021-01-22
上海欧冶供应链有限公司
View PDF0 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The current pricing model reflects the solidification and rigor of the steel spot price system, but it is not flexible and timely enough to respond to market competition and emergencies, and cannot accurately and timely grasp market dynamics and adjust pricing strategies to meet the rationalization needs of the market
The existing steel spot resource pricing method limits rationalized sales in the process of steel destocking transactions

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Steel product spot pricing system and method based on machine learning
  • Steel product spot pricing system and method based on machine learning
  • Steel product spot pricing system and method based on machine learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0101] Example 1: Extract and analyze historical steel spot product delivery and transaction data, and make statistics on packages with large transaction price differences from different dimensions such as production base, grade, tonnage, etc., and divide each interval of different dimensions n divided into There are limited intervals, such as the thickness of hot rolling, and the total upper and lower limits of the interval are 0-1000. After cluster analysis, combined with the contour map, it is divided into (0,1.499), (1.5,1.5), (1.501,1.599) ...(14.501,1000) 45 finite number of intervals. Another example is the feature combination of hot rolling, set up:

[0102] Combined ton weight + place of origin + variety + small variety + actual grade + thickness + width + length,

[0103] Merger of ton weight + place of origin + variety + small variety + merger of actual brand name + thickness + width + length,

[0104] Combination of ton weight + place of origin + variety + smal...

Embodiment 2

[0110] The method of the present invention uses the sklearn package in python to give pricing suggestions for daily transactions, and the data source is all transaction data of historical steel product spot. First use the random forest classification model to predict whether a price increase is needed, and then predict the specific price increase based on the historical characteristics of the data. After 3 months of running the model, the effect is as follows: Accuracy = 70.3% (all predictions are correct (the flexible pricing amount of the package predicted to increase the price is less than or equal to the premium of the actual price increase or the package that is predicted to not increase the price is finally sold without price increase) ) bundles account for the proportion of all transaction bundles); accuracy rate Precision = 90% (that is, the proportion of bundles that are correctly predicted to increase the price to all bundles that are predicted to increase the price);...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a steel product spot pricing method based on machine learning, and the method comprises the steps: screening spot related data of a product according to a business demand, anddetermining an analysis dimension; calculating a characteristic value for each single transaction record; clustering the historical similar orders, and respectively calculating historical order information of the similar orders; the historical order information obtained by the historical order clustering module is utilized to classify the historical orders into price adjustment orders and non-price adjustment orders, and classification model training is performed on data capable of finding historical similar orders and data incapable of finding historical similar orders so as to form two classifiers; respectively calling the two classifiers to predict whether the price is adjusted or not according to whether similar orders can be found or not for the data which is not transacted and is puton the day; for the predicted price adjustment data, calculating a price adjustment amplitude based on historical order information of similar orders in a time period; and optimizing the price adjustment amplitude obtained by the price adjustment amplitude prediction module.

Description

technical field [0001] Technical Field The present invention relates to the field of computer technology, in particular to a machine learning-based spot pricing system and method for steel products. Background technique [0002] At present, when the bundled resources (minimum sales unit) of steel spot are sold on the steel trading e-commerce platform, the resource pricing adopts the marketing pricing method, which focuses on the resource cost structure, marginal rate of return, and market sensitivity to price changes sex. This kind of resource pricing method does not fully refer to the total economic value, and the bundled resource data has many dimensions and complex structure. Relying solely on the experience of business experts, it is impossible to adjust the price strategy in time when responding to market price competition, and cannot comprehensively and accurately assess resources. Price adjustments lead to high or low pricing of resources, reducing the circulation ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06Q30/02G06Q10/04G06K9/62G06N20/10
CPCG06Q30/0206G06Q10/04G06N20/10G06F18/24323G06F18/214
Inventor 王汇丰胡燕张春前王来
Owner 上海欧冶供应链有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com