harshit-bajpai / machine-learning-models

The repository for standard machine learning model developments and related tasks

1. Wine Quality

Exploratory Data Analysis

nb-viewer

This notebook outlines the formats and syntaxes for all the necessary plots in EDA phase of a data (with only numeric and nominal attributes). It also has an elementary GBT model development to validate our variable importance with correlations from EDA.

2. Occupancy Detection

Classification

nb-viewer

This notebook encompasses:

• Data Visualization
  • Histograms
  • Time Series Plots
  • Correlation Matix
  • Pair Plots
• Feature Engineering
• Model Development
  • C - Support Vector Classication
  • Random Forest Classifier
  • Variable Importance

3. Metro Interstate Traffic Volumne

Regression

nb-viewer

This notebook encompasses:

• Data Visualization
  • Histograms
  • Time Series Plots
  • Correlation Matix
  • Box Plots
  • Pair Plots
• Feature Generation
• Model Development
  • Decison Tree Regressor
  • Random Forest Regressor
    • Cross Validation
    • Grid Optimization
  • Variable Importance

4. Concrete Compressive Strength

Regression with interpretability

nb-viewer

This notebook encompasses:

• Data Visualization
  • Histograms
  • Correlation Matix
  • Pair Plots
• Model Development
  • Random Forest Regressor
    • Cross Validation
    • Hyper Paramter Optimization
    • Variable Importance
    • Partial Dependency plots
  • Gradient Boosting Regressor
    • Cross Validation
    • Hyper Paramter Optimization
    • Variable Importance
    • Partial Dependency plots

5. Predictive Maintenance of machines

Xgboost and interpretablity

nb-viewer

This notebook encompasses:

• Data Import
• Data Cleaning and Transformation
• Data Visualization
• Model Development
• Global and local interpretablity through different libaries
  • Using sklearn
  • Using xgboost
  • Using ELI5
  • Using SHAP