YouTube video review: Krish Naik’s LSTM Model for Time Series Forecasting — Deep Learning

Yesterday I found a video that Krish Naik had made concerning univariate time series predictions using Tensorflow. The video was a little over 20 Minutes in length, so I decided to give it a try.

One thing that I found to be a bit awkward is the fact that the data has to be prepared and converted to three columns of data before it can be fed into the LSTM model.

Krish was kind enough to leave the code in a Github repository, which made it easy for me to copy and paste onto my Jupyter Notebook.

I did add a bit to the code Krish made by saving the model, reloading it, and reusing it with another set of code.

Univariate time series analysis

Univariate time series refers to a type of time series analysis where only a single variable is being observed and analysed over time. This variable could be anything, such as stock prices, temperature readings, or the number of visitors to a website.

In the context of TensorFlow, univariate time series analysis involves using the TensorFlow framework to build and train models that can predict future values of the variable being analysed based on past observations. This can be achieved by using various types of neural networks, such…