Khanh Nguyen

Database Structures & Algorithm Data Mining Data Visualization Statistical Analysis Machine Learning Predictive Modeling Driven Data Science Intern ready to thrive in demanding digital intelligence processing environments. Well-informed on latest machine learning advancements. Ready to combine tireless hunger for new skills with desire to exploit cutting-edge data science technology.

Analyzing and Improving Classification Models for Breast Cancer Diagnosis

• Preprocessed and cleaned the data for modeling
• Applied a range of classification models including Logistic Regression, Naive Bayes, SVM, Random Forest
• Bagging classifiers were used to improve the accuracy of the models, particularly for Decision Trees, Random Forest, Logistic Regression, and SVM.
• A voting ensemble model was implemented, which combined the output of the best-performing bagging classifiers and the CatBoost to further boost the overall accuracy of the model to 83%

Linear Regression Model: Using Machine Learning to Predict Taxi Fare Amount

• Used Pandas data frame to resolve missing values, remove unwanted values, and convert categorical columns
• Extracted sample data from a public BigQuery dataset NYC Taxi cab ride using SQL, store the results in Pandas
• Created a Seaborn plot to visualize and explore the data
• Performed linear regression model training and RMSE model evaluation using scikit-learn

Decision Tree Diagnosis: Using Machine Learning to Identify At-Risk Patients.

• Used Pandas data frame to resolve missing values, remove unwanted values, and convert categorical columns to “one-hot encodings”
• Performed model training and evaluation, feature selection to build a machine-learning model for predicting the risk of readmission for patients with heart failure
• Used a decision tree classifier to fit the training data and achieved a testing accuracy of 91.8% and a training accuracy of 94.9%

Data Analysis of Prevalence and Medicare Payment for Chronic Conditions in Four States and National Level

• Cleaned and preprocessed CMS data to remove duplicates, handle missing values, and standardize variables.
• Conducted data analysis using CMS data to determine the prevalence of chronic conditions in the United States and four selected states.
• Analyzed Medicare payments for chronic conditions in the United States and four selected states
• Stratified prevalence rates of chronic conditions by gender and race for the United States and each state.