Skip to content

BharatMaheshwari96/Beverages

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 

Repository files navigation

Beverages

Dataset

The dataset contained in Beverage.csv refers to different physiochemical properties of a beverage produced by a company. This dataset has 4898 observations, with 11 input variables (physiochemical properties) and 1 target variable (quality).
Each row corresponds to a batch of produced beverage. Quality of each batch can be Poor, Normal or Excellent, and one of these labels is assigned by a beverage expert after tasting a sample from a given batch.
The beverage manufacturing company does not want to rely on the beverage expert to label the quality of each batch and wants to build a classification (prediction) model that can predict batch quality on the basis of physiochemical properties.

Task 1

Construct a classification model using random forest in Python. The classification goal is to predict the quality of the produced batch.
Perform all necessary data preparation steps that you may deem fit and choose an appropriate traintest split. Perform necessary hyperparameter tuning to construct an optimal model. Ultimately, arrive at your best random forest model that incorporates a subset of significant features.
Based on information obtained about predictive power of various features used in your final model (using feature_importances), what recommendations can you make to the production department of the company for producing high quality beverage?

Task 2

Implement PCA and K-Means clustering on the given dataset.
PCA Implementation to visualize dataset - How much variance is explained by the first two principal components? Is the resultant plot a good representation of data?
Elbow Plot Creation - What does the elbow plot tell you about the number of clusters (K)? K-Means Implementation - What do you reckon are the correct number of clusters in the dataset?