Skip to content

Predictive modeling project by implementing KNN regression model.

Notifications You must be signed in to change notification settings

weichi21/KNN-Model-Car-Price-Prediction

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 

Repository files navigation

Car Price Prediction by Using K-nearest-neighborhood (KNN) Regression Model

  • Objective: Explore the best numeric feature subset and K-value for car price prediction.

Dataset is from 1985 Ward's Automotive Yearbook
Available from: https://archive.ics.uci.edu/dataset/10/automobile

Here I implemented:

  1. Data Cleansing and Data Transformation
  2. Univariate KNN model with single K-value
  3. Univariate KNN model with multiple K-values (Hyperparameter Tuning)
  4. Multivariate KNN model with single K-Value
  5. Multivariate KNN model with multiple K-Values (Hyperparameter Tuning & K-Fold Cross-Validation)

Conclusion

  • Best Feature Subset: ['city-mpg', 'wheel-base', 'curb-weight', 'highway-mpg', 'peak-rpm']
    Best k Value: 1
    Best Average Accurcy: 86.43%

Future Step

  • With the multivariate KNN hyperparameter tuning, we used f_regression scoring function for best feature subset selection.
    However, f_regression only examines the linear relationship between features and target, and return p-value.
  • We can further apply Mutual Information function from Scikit-learn for feature selection, to see if non-linear results can make the prediction better.
  • We only used numerical features for prediction. It is interesting to explore how combinations of categorical and numeric features ​​can achieve better accuracy by implementing One-Hot Encoding.
pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy