/ML-competition-practice-

Solutions to Machine Learning problems I've worked on

Primary LanguageJupyter Notebook

Machine Learning challenge

(1) Spotify song popularity prediction

  • Participated in a kaggle competition to predict popularity of songs given song features
  • Digged a lot on how to deal with outliers and ML algorithms sensitive to outliers
  • Learned many more techniques to deal Imbalanced data
  • Link to the notebook

(2) Student Acceptance prediction

  • Practiced on a Kaggle Data to predict whether student get admission
  • The Results are:

Logistic Regression:

  • Accuracy of Logistic Regression is 0.475

  • Accuracy of Logistic Regression is 59.5

Support Vector Machine

  • Accuracy of Support Vector Machine is 0.65

  • Accuracy of Support Vector Machine is 50.0

Notes: Here I have not removed the outliers(Need to remove it and train again) Used Label Encoder to handle categorical data(Can use better encoding techniques) Used simple classifiers - Logistic Regression and Support Vector Machine