Skip to content

Multiple models to predict PM2.5 change using the data of Beijing

Notifications You must be signed in to change notification settings

Dalanke/PM2.5_Prediction_Model

Repository files navigation

PM2.5 Prediction Model

6105 Final Project

Using the data of Beijing, we aim at building a model that take some factors into consideration and predict the future concentration of PM2.5

The project first tried HMM and linear regression, then turned into randomforest and ANN for more accurate prediction

Data source:
https://www.kaggle.com/sid321axn/beijing-multisite-airquality-data-set/data

Contribution

Dalanke
kakatcy
majun1997

File Description

The Final version of our work are listed below (please read with order):

  • Intro_and_EDA_start.ipynb: Introduction and EDA
  • HMM-and-Linear-Model.ipynb: HMM and Linear Model
  • RandomForest_final.ipynb: RandomForest Model
  • Tensorflow.ipynb: ANN model

Folders

  • data: raw data and data/cleanup for cleanup data
  • images: images for notebook
  • model: saved trained model
  • paper: research reference

Other Files:

  • simple_clean_up.ipynb: simple clean up function for data clean up
  • Visualization.ipynb: visualization test

Updates

  • April 7, Data set added
  • April 8, EDA for Shunyi data
  • April 9, EDA for Shunyi(cont'd) and cleanup for other data set
  • April 14, Theoretical support/papers/images
  • April 14, Bayesian inference update
  • April 17, More linear model
  • April 20, Add HMM, final version for HMM-Linear-Model

About

Multiple models to predict PM2.5 change using the data of Beijing

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published