The Combined Forecasting Model For Time Series

In this work, I want to reproduce my bachelor thesis work into a Python version and describe it in English

Objectives

This work wants to improve the accuracy of housing price prediction using a combined model that integrates the advantages of both ARIMA and NN models.

Method

1. Univariate Time-series prediction via Neural Network (NN) in TensorFlow.
2. Univariate Time-series prediction via Autoregressive integrated moving average (ARIMA) in statsmodels.
3. Combining these two models into one using the inverse-variance weighting method.

Dataset

https://www.kaggle.com/datasets/yasserh/housing-prices-dataset?select=Housing.csv

Code

programming

Conclusion