Gold Price Prediction Model

This project involves the development of a gold price prediction model using a combination of Random Forest and ARIMA models. The model was trained on historical gold price data and is capable of forecasting day-wise gold prices with 97% accuracy, significantly improving from an initial accuracy of 76%.

Project Overview

We developed a gold price prediction model using time series analysis, focusing on improving forecasting accuracy. The model leverages advanced techniques like Random Forest and ARIMA to predict gold prices on a daily basis. The project aims to provide accurate predictions for investors, traders, and analysts in the precious metals market.

Key Features

• Improved Prediction Accuracy: Achieved an impressive 97% accuracy after applying Random Forest and ARIMA models, up from 76% in initial experiments.
• Data Preprocessing: Performed extensive data cleaning, manipulation, and exploratory data analysis (EDA) on time series data spanning from 2016 to 2021.
• Forecasting: Deployed the prediction model using Streamlit, making it accessible for real-time day-wise gold price forecasting.

Technical Details

• Data Source: Time series data (2016-2021) on historical gold prices.
• Tools Used:
  • Python for data processing, modeling, and analysis
  • Random Forest for machine learning-based predictions
  • ARIMA for time series forecasting
  • Streamlit for model deployment

Steps Taken

1. Data Cleaning: Cleaned the raw data to handle missing values, outliers, and inconsistencies.
2. Exploratory Data Analysis (EDA): Visualized and understood the data trends, patterns, and seasonal variations.
3. Modeling:
   • Built Random Forest model for price prediction.
   • Applied ARIMA model for capturing time-dependent patterns.
4. Model Improvement: Enhanced accuracy from 76% to 97% through hyperparameter tuning and model refinement.
5. Deployment: Used Streamlit to deploy the final model, making it accessible for real-time predictions.

Installation

To run the project locally, follow these steps:

1. Clone the repository:

   git clone https://github.com/your-username/gold-price-prediction.git
   cd gold-price-prediction

2. Install the required dependencies:

   pip install -r requirements.txt

3. Run the Streamlit app to interact with the model:

   streamlit run app.py

Project Structure

gold-price-prediction/
│
├── data/                   # Contains raw and processed data
│   └── gold_prices.csv     # Gold price data from 2016 to 2021
│
├── models/                 # Contains model scripts
│   ├── random_forest.py    # Random Forest model code
│   ├── arima_model.py      # ARIMA model code
│
├── app.py                 # Streamlit app for deployment
├── requirements.txt       # Project dependencies
└── README.md              # Project documentation

Future Improvements

• Incorporate additional features such as geopolitical events, market trends, and currency fluctuations.
• Experiment with other models like LSTM for deep learning-based time series forecasting.