Routing in a Sparse Graph: a Distributed Q-Learning Approach

What’s Happening

Not gonna lie, Distributed agents need only decide one move ahead.

The post Routing in a Sparse Graph: a Distributed Q-Learning Approach appeared first on Towards Data Science. Youve probably heard about the Small-World Experiment , conducted in the 1960s. (wild, right?)

He devised an experiment by which a letter was given to a volunteer person in the United States, with the instruction to forward the letter to their personal contact most likely to know another person the target in the same country.

The Details

In turn, the person receiving the letter would be asked to forward the letter again until the target person was reached. Although most of the letters never reached the target person, among those that did (survivorship bias at play here!

), the average number of hops was around 6. The six degrees of separation has become a cultural reference to the tight interconnectivity of society.

Why This Matters

I am still amazed that people with ~10 2 contacts manage to connect with a random person in a network of ~10 8 people, through a few hops. Lets assume I am asked to send a letter to a target person in Finland 1 . Unfortunately, I dont have any contacts in Finland.

As AI capabilities expand, we’re seeing more announcements like this reshape the industry.

Key Takeaways

• On the other hand, I know someone who lived in Sweden for many years.
• Perhaps she knows people in Finland.
• If not, she probably still has contacts in Sweden, which is a neighboring country.
• She is my best bet to get closer to the target person.

The Bottom Line

This problem of transmitting the message in the right direction offers an opportunity to have fun with ML! Nodes do not perceive the whole network topology.

Thoughts? Drop them below.