How to Survive Christmas as a Data Scientist | by Isabelle Flückiger | Dec, 2020
1. Search for some data to plot stem-leaf diagrams in the form of a Christmas tree
Nice data for descriptive plots are always a pleasure. And during Christmas time, the data can also have Christmas tree or star forms.
So, try to find some nice data, do some descriptive analytics, print it, and hang your graph on the wall, or start a social media challenge about the best plots with your friends.
2. Fit some neural networks to predict the minutes until the next time “Last Christmas” from Wham! is played on the radio
The title says it all.
We all know that song. Some love it, and many hate it, including me. But it belongs to Christmas like regression to the data science work.
How often the song is played depends on the country you live in and the radio station you listen to. On the homepage of World Airplay Radio Monitor, you can find some statistics about Christmas songs on radios. And it is really often. By scanning through the radio station data, I discovered that the song was played up to 24 times per day, or on average every hour.
So, choose your favorite radio station and predict!
Should you have a result, please post it in the comments.
3. Develop a classifier for reindeer
We all know Rudolph the red-nosed reindeer.
Why not take the challenge to build a Rudolph the reindeer classifier?
4. Bring Santa Claus and Rudolph to the cloud
Well, that is a very delicate task. It begins with the question about which cloud provider? Or should we set up a multi-cloud environment?
Let’s share the experience in the comment field.
5. Create your Christmas Story with some natural language generation algorithm
Once upon a time, there lived at the North Pole, young reindeer named Rudolph. He was a lively, happy reindeer and loved playing in the snow. But one day, the other reindeer noticed that he had a very large, bright red nose, and they teased him about it without mercy.
[generated] When Rudolph was about six years old, the other reindeer became afraid of his nose and refused to play with him. They then decided to get rid of him by making him a snowman. But Rudolph had other ideas and decided that he wanted to be a bear instead of being a snowman. Now, Rudolph was in charge of building a big house for his mother and brothers. The house was enormous, and it was decorated in bright colors. But Rudolph was not happy being a bear and decided to become a fox. Rudolph was delighted, and now he had a new life.
Well, I assume you can do that better…
Please post your generated story in the comments.
6. Predictive Maintenance for Santa Claus’ sleigh
Collect sleigh data, predict the next breakdown, and optimize the maintenance work while Santa Claus brings all the children’s presents.
7. Develop facial recognition for the Santa Claus
This could look then like this:
Please mention your accuracy in the comments.
8. Predict the weight of the presents on the sleigh with linear regression
Based on the sleigh’s speed and flight altitude, fit a regression model to predict the weight of the presents on the sleight.
9. Optimize the path of Santa Claus in your region/country
Bringing all the presents as efficiently as possible to all the small and large children is an enormous effort. You can support by optimizing the transport route in your country and provide guidance to Santa Claus.
On the 24th, you find the path chosen by Santa Claus on the NORAD tracker:
10. Build a robot reindeer
And last but not least, why not enhance a Lego kit to a real working robot and train it with neural networks?
Read More …