IBM Advanced Data Science Capstone Project - Time Series Prediction

Time Series Prediction using LSTM and SARIMAX

Table of Contents

1. Installation
2. Project Motivation
3. File Descriptions
4. Results
5. Licensing, Authors, and Acknowledgements

Installation

In order to run the code for the LSTM Model, Keras needs to be installed. This aside, no other necessary libraries should be required to run the code. The code should work with no issues using Python versions 3.*.

Project Motivation

As part of the IBM Advanced Data Science Capstone Project, I tried to forecast the Website Traffic of a News Website in Germany comparing an LSTM with SARIMAX model.

1. Collecting the data & filtering out only the required data
2. Analysing the data (checking for anomaly or missing values)
3. Preparing the data for the models
4. Running the models and tuning the hyper parameters
5. Gathering, evaluating and discussing the outcome

The dataset is not publicly available due to the ownership of the data. However, you can see pieces of the analysis and evaluation here.

File Descriptions

There are 4 notebooks available here to show the above steps that I went through.

For the final project, I had to held a presentation as well as uploading a video of the presentation, which could be watched here

Results

The main findings of the code can be found in the 4th Chapter of the juypter notebooks here.

Licensing, Authors, Acknowledgements

Must give credit to the company for the data. Please, feel free to use the code here as you like!