Airline Passengers Prediction Using My Stochastic Time Series Algorithm From Scratch in Python

My Stochastic Time Series Algorithm — Ashutosh Malgaonkar

Get an email whenever Ashutosh Malgaonkar publishes.

Github link to my code:

https://github.com/ashm212/Stochastic-Time-Series/blob/122ef2e5ce861d3a97db6be633799317be91e892/Stochastic_Time_Series.ipynb

I. Accessing Data From Kaggle using the Kaggle API

II. Algorithm

a. Input to dataframe

b. Programming the Magnitude

c. Programming the Direction Probability

d. Programming the Stochastic Prediction

VI. Graphing ( https://www.geeksforgeeks.org/how-to-plot-timeseries-based-charts-using-pandas/)

VII. MAPE

I. Accessing Data From Kaggle

I accessed the data using a Kaggle API. This part is code from StackOverflow from an individual who shared the general steps on how to access Kaggle data. I am citing the link to his code, but I will also walk through it in this article the details. These steps are fairly normal for any dataset.

https://colab.research.google.com/drive/1DofKEdQYaXmDWBzuResXWWvxhLgDeVyl

Now, let us do a walkthrough of how it happened for this project:

First, go to kaggle.com and go to your account page. There scroll down till you get to the API section, which looks like the picture below. Click “Create New API Token”. This will download a file to your computer called kaggle.json.

Next, go to your collab notebook:

from google.colab import files
files.upload()

Run the above codeblock. Here is a picture as well: