/sales-prediction

Primary LanguageJupyter Notebook

Sales Prediction Using SARIMA and Deep Learning Models

This project aims to forecast sales data using a combination of traditional time series models and deep learning approaches. Specifically, we use SARIMA for capturing seasonality and trends, and deep learning models (LSTM, GRU, RNN) for learning complex temporal patterns. An ensemble of SARIMA and LSTM predictions is also explored to enhance forecasting accuracy.

Project Structure

  • code/: Contains the sales data file and Jupyter notebooks used for experimentation and EDA.
  • models/: Saved models and scripts for training and evaluation with generated plots for loss functions and prediction results.
  • dataset: univariant time-series sales dataset for analysis.
  • README.md: Project overview and instructions.

Requirements

  • Python 3.x
  • TensorFlow
  • Keras
  • Pandas
  • Numpy
  • Matplotlib
  • Statsmodels
  • Scikit-learn