gabrielecola / BayesianApproach_CO2

Bayesian Approach to contrast C02 Emission

For the report click 👉 here

Table of Contents

• C02 Emission count prediction using Bayesian Model
• 1. Problem Statement
• 2. Data Description
  • Attribute Information
    • Inputs
    • Output
• 3. EDA
• 4. Modelling Evaluation
• 5. Results

C02 Emission count prediction using Bayesian Model

This dataset captures the details of how CO2 emissions by a vehicle can vary with the different features. The dataset has been taken from Canada Government official open data website.This contains data over a period of 7 years. Thus, using Bayesian model to predict the count of C02 emission could be useful in solve this dangerous problem

1. Problem Statement

Predicting the count of C02 emission given some qualitative and quantitative attribute.

2. Data Description

Data is obtained from Kaggle

• Number of instances - 7385

• Number of attributes - 12

  Attribute Information

  Inputs

  • make
  • model
  • vehicle_class
  • engine_size_l
  • cylinders
  • transmission
  • fuel_type
  • fuel_consumption_city_l_100_km
  • fuel_consumption_hwy_l_100_km
  • fuel_consumption_comb_l_100_km
  • fuel_consumption_comb_mpg
  • co2_emissions_g_km

  Output

  • co2_emissions_g_km

  3. EDA

  

  4. Modelling Evaluation

  • Algorithms used
    • Bayesian Linear Regression (1 Approach)
    • Bayesian Linear Regression (2 Approach
    • Bayesian Poisson Regression
    • Bayesian Poisson Overdispersion Regression
    • Bayesian Hierarchical Poisson model
  • Metrics: RMSE (Root Mean Squared Error)

  5. Results