Machine Learning Web Platform

Overview

This web application allows users to upload datasets, train machine learning models, and visualize model performance with metrics and plots. It supports both classification and regression tasks and provides an interactive dashboard for managing datasets and models.

---

System Requirements

• Python Version: 3.10+

---

Environment Setup

1. Clone the Repository

   git clone https://github.com/gdjones-526/CS456-Project.git
   cd CS456-Project
   

2. Setup Environment

   python -m venv venv
   # Windows
   venv\Scripts\activate
   # macOS/Linux
   source venv/bin/activate
   

3. Install requirements and migrate database

   pip install -r requirements.txt
   python manage.py migrate
   

4. Run Server

   python manage.py runserver
   

This guide provides a step-by-step walkthrough to use the website.

---

1. Account Login / Sign Up

1. On first visit, you will see the Account Login Page.
2. Enter your username and password if you already have an account.
3. If you do not have an account, click Sign Up Now.
4. On the Sign Up Page:
   • Create a username.
   • Create a password that is unique from your username and at least 8 characters long.
   • Confirm your password.
   • Click Sign Up.
5. After signing up, you will be automatically logged in.

---

2. Dashboard Overview

After logging in, you will see the Dashboard Screen.

A. Data Ingestion and Pre-Processing

1. Under the Data Ingestion and Pre-Processing tab:

   • Upload your dataset (CSV, Excel, or TXT format).
   • Optionally, add a description for your dataset.
   • Click Upload Dataset.

2. You will be redirected to the Data Set Dashboard:

   • View dataset features: total rows, columns, file size, and missing values.

---

B. Training a Model

1. Scroll down to the Actions tab and click Train Model.
2. In the Train Machine Learning Model View:
   • Select your target variable from the dropdown.
   • Select features that the model will use.
   • Choose the task type: Classification or Regression.
   • Based on the task type, select the algorithm for training.
   • Under Model Configuration:
     • Enter a Model Name.
     • Optionally adjust:
       • Test Size
       • Validation Size
       • Missing Value Strategy
       • Random State
   • Click Start Training.
   • Training may take a few seconds.

---

C. Viewing Model Results

1. After training, the Model View displays:
   • Classification Models: Accuracy, Precision, Recall, F1 Score
   • Regression Models: Error Metrics, Accuracy, MSE, RMSE
2. Scroll down to see:
   • Training Configuration
   • Model Configuration
   • Features Used
   • Visualizations: ROC Curve, Confusion Matrix, Feature Importance

---

3. Managing Data Sets

1. From the Dashboard, click ML Platform in the top-left corner.
2. Scroll to Your Data Sets:
   • View uploaded datasets, descriptions, upload date, and processing status.
   • Delete a dataset if no longer needed.

---

4. Model Evaluation and Comparison

1. Under Model Evaluation and Metrics:

   • Models are categorized by type: Classification or Regression.
   • Models are sorted by accuracy (highest to lowest).

2. Comparing Model Figures:

   • Select Model 1 and Model 2 from dropdowns.
   • Choose the Figure Type: ROC Curve, Confusion Matrix, or Feature Importance.
   • View side-by-side visualizations to compare models.

---