umangu Goto Github PK

• 😄 Pronouns: He/Him/His
• 🔭 I’m currently working on: a multivariate time-series forecasting modelling project
• ❤️ My favorite language: SQL
• 🌱 I’m currently learning: Recurrent nueral nets
• 💼 I’m always looking to collaborate with: people from any field
• 💬 Ask me about: anything, I am happy to help
• 🌍 I support: KoBo Toolbox, Child's Play, Intelehealth, Code For Progress
• 💜 Interests: philosophy, travel, olympic lifting, internet culture, video games

• 👾  
• 🌐  
• ⚙️  
• 💻  

Implementing Feed-Forward neural networks on the "Wool" dataset while looping through 1-7 hidden units and changing the train/test ratio.
Libraries Utilized: neuralnet

SQL Stored Procedures with R's external scripts to predict a client's probability of defaulting on loan using support vector machines.
R-based web application that allows a user to upload a dataset and get the prediction on client's likelihood of defaulting on loan.
Libraries Utilized: readxl, e1071, shiny, shinydashboard, shinythemes

Analyzing and comparing supervised-classification models on the "Appointment1000" dataset.
Libraries utilized: MASS, class, caTools