An analysis project which includes data visualization/EDA analysis and machine learning method(logistic regression,KNN,Random Forest and GBM)to build a binary classification model to predict if individuals will leave the organization.
An analysis project mainly focus on using streamlit to create a analysis dashboard to analyze the motor vehicle collisions in New York City. It includes prepoccessing data and data analysis 2 parts.
An analysis project using Bayesian methods to find the posterior distribution of the coefficients.We will try to discover how different priors on β will affect the posterior distribution.We suggest three priors, a flat prior, a normal prior with Inverse-Gamma variance, and a normal prior with Half-Cauchy vairance.
A CNN classification project by using Resnet50 and VGG16 model to predict if the patients' X-Ray image is a normal, viral pneumonia or COVID-19 image.
A NLP classification project using GRU,Bi-LSTM and BERT model to identify the offensive language on twitter.This project is based on the SEMEVAL 2019 TASK6. From the results we propose BERT as our recommended approach, given its higher macro F1-score of 81.73 percent and accuracy of 86.18 percent.