Evalución frecuenta Trabajo en ejercicio en R studio y Presentaciones.

Three-model comparison for predicting loan approval outcomes (4,269 records). Covers logistic regression, decision tree, and random forest — each evaluated with a confusion matrix, feature importance plot, and ROC/AUC score. Built as a reusable modeling template for datathon competition prep.

Dokumen ini berisi visualisasi data kategorik dan data numerik perdagangan antar wilayah Indonesia setiap provnsi.

Visual exploration of a loan approval dataset (4,269 records) covering target variable balance, numeric and categorical distributions, correlation heatmap, and bivariate analysis of key predictors against loan outcome. Built as a reusable EDA template for datathon competition prep.

Data cleaning workflow for a loan approval dataset (4,269 records). Covers column standardization, type conversion, missing value audit, outlier capping, and feature engineering. Built as a reusable template for datathon competition prep.