Air Quality Prediction of Relative Humidity - Regression
Jupyter Notebook 100.00%
air-quality-prediction's Introduction
Air Quality Prediction of Relative Humidity - Regression
The dataset contains 9358 instances of hourly averaged responses from an array of 5 metal oxide chemical sensors embedded in an Air Quality Chemical Multisensor Device. The device was located on the field in a significantly polluted area, at road level. Data were recorded for a period of one year (March 2004 - February 2005) in an Italian city. (Dataset)
Features
Date
Date (DD/MM/YYYY)
Time
Time (HH.MM.SS)
CO(GT)
True hourly averaged concentration CO in mg/m^3 (reference analyzer)
PT08.S1(CO)
PT08.S1 (tin oxide) hourly averaged sensor response (nominally CO targeted)
NMHC(GT)
Non Metanic HydroCarbons concentration in microg/m^3 (reference analyzer)
C6H6(GT)
True hourly averaged Benzene concentration in microg/m^3 (reference analyzer)
Random Forests and Polynomial Regression did a great job predicting relative humidity.
Since the default parameters already did a great job, there was no room left for improvement of the models with the help of grid search.
Based on the correlation between the variables and the target, removal of less correlated variables did not improve the accuracy of the prediction, therefore, all the variables were used in the prediction.
In addition, for Linear and Polynomial regressions not many parameters can be changed, therefore, grid search contribution to the improvement of the model score was very low.