Skip to content

Krish-024/predictive-analytics-app

Repository files navigation

PredictiTrend Analytics 📈

PredictiTrend is a powerful, web-based predictive modeling platform designed to help you transform historical data into actionable future insights. Whether you're tracking retail sales, stock market fluctuations, or IoT sensor data, PredictiTrend provides the tools to clean, model, and visualize your data with ease.

🚀 Key Features

  • Intuitive Data Ingestion: Drag-and-drop CSV uploader with automatic column and type detection.
  • Advanced Forecasting Models:
    • Linear Regression: Ideal for identifying long-term growth or decline trends.
    • Simple Moving Average (SMA): Perfect for smoothing volatile data and spotting cyclical patterns.
  • Interactive Visualizations: High-fidelity charts powered by Recharts that overlay historical facts with projected trends.
  • Statistical Accuracy Metrics: Real-time calculation of R-Squared (Goodness of Fit) and Mean Absolute Error (MAE) to evaluate model reliability.
  • AI-Powered Insights: Integrated Gemini AI that analyzes your statistical results and provides executive summaries of data trends.
  • Data Export: Download your combined historical and predicted datasets as a single CSV file for further analysis.

🛠️ Tech Stack

  • Frontend: React 19, TypeScript, Vite
  • Styling: Tailwind CSS 4
  • Animations: Motion (formerly Framer Motion)
  • Charts: Recharts
  • Mathematics: Simple-Statistics
  • Data Parsing: PapaParse
  • AI Engine: Google Gemini API (@google/genai)

🏁 Getting Started

Prerequisites

Installation

  1. Clone the repository:

    git clone https://github.com/your-username/predictive-analytics-app.git
    cd predictive-analytics-app
  2. Install dependencies:

    npm install
  3. Set up environment variables: Create a .env file in the root directory and add your Gemini API key:

    VITE_GEMINI_API_KEY=your_api_key_here
  4. Run the development server:

    npm run dev
  5. Build for production:

    npm run build

📊 Usage

  1. Explore samples: Click the "Explore with Sample Data" cards on the landing page to see an instant demonstration.
  2. Upload your own: Drag a CSV file into the uploader.
  3. Configure axes: Pick your Time (X-Axis) and Value (Y-Axis) columns.
  4. Choose a model: Select the regression or moving average model that best fits your data.
  5. Analyze: Click "Run Model Analysis" to generate the forecast and AI insights.

📄 License

This project is licensed under the Apache-2.0 License.

About

Here’s a **~320-character version**: **PredictiTrend Analytics** is a web-based platform that transforms historical data into actionable insights using Data Science and Machine Learning. It supports forecasting models like Linear Regression, interactive visualizations, and AI-powered insights to analyze trends and predict future outcomes.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages