MSBA-Capstone-MasterControl-GroupProject

📊 MSBA Capstone - Group Project Dashboard

MSBA IS 6813 | Spring 2026

---

🛠️ Functional Hub & Assignments

Core Tools 🛠️	📋 Specs	📊 Data Room	📁 Shared Google Drive	🌐 Group Dashboard	💻 GitHub Repo
Assignments 📂	📄 01 Business Problem Statement	📂 02 EDA	📂 03 Modeling	📂 04 Presentation	🚫

---

📅 Deadlines

Phase	Milestone	Hard Deadline
🟢	Business Problem Statement	Jan 28
🟡	EDA Group Notebook	Feb 18
⚪	Modeling Notebook	Mar 18
⚪	Practice Presentation	Apr 05
⚪	Final Sponsor Delivery	Apr 08/15
⚪	Portfolio & Peer Eval	Apr 19

---

⚙️ Notebook tips

1. Notebook Standards & The “Golden” YAML

Primary Directive: Copy this block exactly into the top of every .qmd file.

---
title: 
subtitle: 
date: "Spring 2026"
format:
  html:
    theme: journal
    toc: true
    toc-depth: 3
    toc-float: true
    number-sections: false
    code-fold: true
    code-tools: true
    df-print: paged
    highlight-style: github
  pdf:
    documentclass: article
    geometry:
      - margin=1in
    toc: true
    number-sections: false
    colorlinks: true
    mainfont: "Arial"
    sansfont: "Arial"
    monofont: "Courier New"
editor: visual
---

2. Standard Setup & Parallel Processing (Copy-Paste)

Rule: Use these blocks to initialize your environment. They include Dynamic Core Selection to maximize performance on any machine without crashing it (N-1 logic).

🟢 R Setup (Tidyverse + Parallel)

# Load Core Packages
if (!require("pacman")) install.packages("pacman")
pacman::p_load(tidyverse, here, parallel, doParallel)

# Dynamic Parallel Processing (Detects your hardware)
# Leaves 1 core free for the OS to prevent freezing
num_cores <- parallel::detectCores(logical = FALSE)
cl <- makeCluster(num_cores - 1)
registerDoParallel(cl)

print(paste("Cluster active with", num_cores - 1, "cores."))

🔵 Python Setup (Pandas + Multiprocessing)

import pandas as pd
import numpy as np
import multiprocessing
from pyprojroot import here

# Dynamic Core Selector
# Use 'n_jobs' in Scikit-Learn models (e.g., n_jobs=n_jobs)
n_jobs = multiprocessing.cpu_count() - 1

print(f"Parallel processing enabled: {n_jobs} cores available.")

⚡ Performance Note: Why Parallel Processing?

Standard R and Python scripts run linearly on a single CPU core, leaving 80-90% of your computer’s power idle. By enabling parallel processing (as shown above), we distribute computations across multiple cores simultaneously.

• Impact: This is critical for resource-intensive tasks like Random Forest, XGBoost, or Grid Search.
• Result: It can reduce model training time by 4x to 8x depending on your hardware.

3. Foolproof Data Loading (Polyglot Paths)

Try not to use absolute paths (e.g., C:/Users/Thomas/...).

For R (using here):

library(here)
# Automatically finds the project root (where .git is)
df <- read.csv(here::here("data", "application_train.csv"))

For Python (using pyprojroot):

from pyprojroot import here
# Automatically finds the project root
path = here("data/application_train.csv")
df = pd.read_csv(path)

---

🧠 Repository Architecture & Usage Flow

Visual map of how files, data, and code interact within this repository.

┌─────────────────┐       ┌────────────────────┐      ┌──────────────────┐
│  📂 data/       │       │  📂 notebooks/     │      │  📂 output/      │
│  (Local Only)   │──────▶│  (Code Execution)  │─────▶│  (Deliverables)  │
│  Raw .csv Files │       │  .qmd Analysis     │      │  .csv / .png     │
└─────────────────┘       └────────────────────┘      └──────────────────┘
        │                           ▲
        │                           │
        └────── (Load via 'here') ──┘

📂 Physical Directory Structure

├── data/               # RAW data (Local only - Git ignored)
├── notebooks/
│   ├── 01_Business_Problem/
│   ├── 02_EDA/
│   ├── 03_Modeling/
│   ├── 04_Presentation/
│   └── individual/     # Individual "Sandboxes" for portfolio
├── output/             # Exported .csv results and .png plots
├── docs/               # Meeting notes and sponsor requirements
└── README.md           # This Hub

---

📞 Contact Information

Team Member	Email (Personal)	Email (University)	Phone
Thomas Beck	thomasscottbeck@gmail.com	u0399590@utah.edu	+1 (801) 631-2080
Max Ridgeway	[TBD]	u1230181@utah.edu	+1 (801) 597-3824
Astha KC	asthakc.us@gmail.com	u1561947@utah.edu	+1 (971) 500-6757

---

Note: Before starting any work session, run git pull to sync the latest model changes from the team.