/Fraudulent-Job-Post-Prediction

Based on Job title, description and requirements, a given job entry is classified into REAL/FAKE using 7 ML models.

Primary LanguageJupyter Notebook

Predicting Fraudulency based on Job Posts

  1. Visualizing Missing Values
  2. WORDCLOUD on Job Titles
  3. Splitting Locations into Country, State & City
  4. Converting salary ranges into Min & Max
  5. Label Encoding Categorical Features
  6. Cleaning Text Features by removing STOPWORDS and Lemmatizing Words using NLP
  7. OVERSAMPLING Target Variable
  8. Scaling Data using MINMAXSCALER
  9. Plotting AUC and Accuracies of following Models:
    • Logistic Regression
    • Support Vector Classifier
    • MultiLayer Perceptron Classifier
    • KNN Classifier
    • Decision Tree Classifier
    • XGBoost Classifier
    • Random Forest Classifier