TRYEngineering - AI Adventures - 11

In several ways, a Rubik's Cube can serve as an effective analogy
for machine learning:
* Complexity and Patterns: Just like solving a Rubik's
Cube involves recognizing patterns and applying
algorithms, machine learning involves identifying
patterns in data and applying algorithms to make
predictions or decisions.
* Step-by-Step Process: Solving a Rubik's Cube
requires a series of steps to reach the solution, similar
to how machine learning models go through a series of
steps (data preprocessing, training, evaluation) to learn
and make accurate predictions.
* Optimization: Both involve optimization. In a Rubik's
Cube, you aim to find the most efficient way to solve it.
In machine learning, you aim to optimize the model's
performance by tuning hyperparameters and
improving accuracy.
* Learning from Mistakes: When solving a Rubik's
Cube, you often learn from mistakes and adjust
your approach. Similarly, machine learning models
learn from errors during training and adjust to
improve performance.
ML is the Main
Approach.
Machine learning, which uses examples (data) to learn
how to perform a task, is currently the main method
by which AI learns, eventually becoming capable of
performing the task on its own in a new situation.
Machine learning teaches AI in three main ways:
supervised learning, unsupervised learning, and semisupervised
learning.
* Supervised learning uses examples that have the
right answers attached to them to help train the AI. This
enables the AI to recognize patterns in its input and
make predictions from previously unseen data.
* Unsupervised learning uses examples without the
right answers. Developers may use this method to save
time or money or because they don't know the correct
answers for the data, but when AI learns this way, it
discovers its own patterns from the examples, which it
can apply to new data when it's ready.
* Semi-supervised learning is a mix of supervised and
unsupervised learning. The machine learns from both
the correct and incorrect answers in some examples.
Depending on the goal of the AI tool, one of these types
of machine learning might be preferred, but they can all
solve some of the same problems, like detecting faces
or facial features in photos!
However, there
are differences
too. Machine
Learning often
deals with
probabilistic
outcomes and
large datasets,
whereas a Rubik's
Cube has a finite
number of states and
deterministic solutions.
Overall, the Rubik's Cube can serve as a helpful analogy
to explain the iterative and pattern-recognition aspects of
machine learning.
WATCH: The Daisy Video by Rubik's & MAKE:
The Daisy on a Rubik's cube.
https://www.youtube.com/watch?app=desktop&v=2mgtMpoAd8c
TRYEngineering - AI Adventures

Table of Contents for the Digital Edition of TRYEngineering - AI Adventures
