Projects
Here are the few key projects on which I worked upon during my learning journey
Patient survival in hospital
Using the dataset of Greenland's hospital patients I have analysed the features using feature selection methods RFE, Boruta and achieved the best results using xgboost to predict whether a patient will survive or not.
This analysis will help hospital to analyse the main factor for the death of majority of the patients.
![](https://static.wixstatic.com/media/11062b_16e953250db143b99693621c098d0bd6~mv2.jpg/v1/fill/w_345,h_232,al_c,q_80,usm_0.66_1.00_0.01,enc_auto/11062b_16e953250db143b99693621c098d0bd6~mv2.jpg)
Twitter sentiments analysis
After doing data pre-processing and
exploratory data analysis I have used
the bag of words and word embedding
model to train a model and achieved
best results with hyperparameter tuning
using xgboost algorithm.
![](https://static.wixstatic.com/media/5a3e0ec71e8f4781a67cde702803d352.jpg/v1/fill/w_347,h_231,al_c,q_80,usm_0.66_1.00_0.01,enc_auto/5a3e0ec71e8f4781a67cde702803d352.jpg)
Predicting employee attrition
Using IBM employee dataset , after standardising and normalising the data into same format using label and one-hot encoding Random forest classifier algorithm gave the best results on test data and gave the probability for chances of a employer to be kick out of the organisation.
© 2020 by Rachit Agarwal
Proudly created with Wix.com