You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I worked on a dataset of covid vaccination which includes informaiton about name of countries, name of vaccines and different statistics about vaccinations in each date.
The focus of this project is on Data Cleaning, Data Analysis, Data Extraction for story-telling, Gaining an insight from Pycaret Regression tool over best ML algorithms,and finally comparing 4 top models for selecting the best.
Fisrt, different raws of raw dataset is investigated, to gain as much idea for storytelling, as possible. In this direction, interesting plots and tables are extracted.
After choosing the country for modelling, data is cleaned. For filling missing values, a linear regression approach is chosen, based on the available data for this country in other dates.
Then outliers are removed and data is scaled.
After applying Pycaret over the cleaned data, top models are selected for hyper-parameter tuning:
1. RandomForest
2. ExtraTrees
3. GradientBoosting
4. CatBoost
Then the performance of these models are compared together based on MAE factor.
The best model (ExtraTrees model) is then saved using joblib module for further use.
About
The focus of this project is on Data Cleaning, Data Analysis, Data Extraction for story-telling, Gaining an insight from Pycaret Regression tool over best ML algorithms,and finally comparing 4 top models for selecting the best.