/Wine-Quality-Test

This is my handson ML project. Here I will went through a paper to know how wine quality can be tested, alongside will make presentations and video demonstration for this.

Primary LanguageJupyter Notebook

Wine Quality Test


Initials

Wine

  • an alcoholic drink typically made from fermented grapes
  • once viewed as a luxury good
  • nowadays wine is increasingly enjoyed by a wider range of consumers

Idea Overview

  • Build an interface to predict the quality of the red wine. (Through data mining approach)
  • Result of the system: Quality of wine, given chemical information and machine learning model.

Motivations

Wine Itself!

  • Lowers bad cholesterol
  • Keeps heart healthy
  • Regulates blood sugar
  • Reduces the risk of cancer and many more!

Wine Industry

To support its growth, wine industry is investing in new technologies for both wine making and selling processes. Industry desition takers must check the followings

Wine Certification

  • prevents the illegal adulteration of wines (to safeguard human health)
  • assures quality for the wine market

Quality Assessment

  • often part of the certification process
  • used to improve wine making (by identifying the most influential factors)
  • used to stratify wines such as premium brands (useful for setting prices)

Existing Approach Limitations

Decisions on wine quality prediction are mostly done scarce and considers small datasets.

  • 1991’s the “Wine” dataset includes 178 examples with measurements of 13 chemical constituents
  • 1997’s the “Wine” dataset includes 170 samples from Germany but predict 100% accurately
  • 2001’s wine dataset includes only 36 examples were used and 6% error achieved

Contributions

  • Proposed a data mining approach to predict human wine taste preferences that is based on easily available analytical tests
  • Compared to other domain, a large dataset is considered with white and red vinho verde samples from northwest Portugal have increased by 36% from 1997 to 2007

Why Portugal's Dataset?

  • top ten wine exporting country, with 3.17% of the market share in 2005
  • Exports of its vinho verde wine (from the northwest region) have increased by 36

Advantages of This Project

  • Such model is useful to support the oenologist wine tasting evaluations and improve wine production
  • similar techniques can help in target marketing by modeling consumer tastes from niche markets

Project Demonestration

  • Actual paper link is given here
  • Dataset Link is given here
  • Project Github Link is given here

Contributor