ELVIS TATA TANGHANG

Trained and evaluated four machine learning models to select the best model for the Titanic survival prediction. Turned the hyperparameters using grid search and selected the best model based on the cross-validation accuracy. Validation accuracies; decision tree (91.8%), KNN (86.5%), logistic regression (83.4%) and decision tree (76.05%)

Analyzed and interpreted complex data; weather and electricity consumption in France 2017. Generating insights and developing data driven solutions Trained and optimized an OLS model through stepwise feature selection with 88.2% accuracy. Virtualized the relationship between weather features and the daily energy consumption.

Prepared and explored customer’s credit card data to build a dual machine learning model for effective services. Developed and evaluated an unsupervised ML model to segment customers. Trained a decision tree model with 96.2 % accuracy and deployed on a webpage using Flask.