Get a high-level review of artificial intelligence, unsupervised machine learning, and supervised machine learning.
- [Instructor] I want to begin by reviewing some terms that will allow us to identify where we'll be focused in this course. My favorite definition of artificial intelligence is one that I'm going to borrow from a thought leader in this space named Colin Shearer. "A computer doing something that, "if done by a human being, "would be judged to be intelligent." Now of course, artificial intelligence is a huge field and involves all kinds of things like visual recognition, driverless cars, and devices in our home, like smart thermostats or devices like Amazon Alexa.
We're going to be focused on traditional machine learning, which encompasses two of the most important topics of all, supervised and unsupervised machine learning. I'm going to define traditional machine learning in the following way, a broad term that generally refers to presenting carefully curated data to computer algorithms that find patterns and systematically generate models. These models are going to be in the form of formulas or rule sets.
While the algorithms are explicitly programmed by humans, the models are automatically generated. We're going to be even more focused than that. Throughout the course, we're going to be talking about supervised machine learning. Supervised machine learning is going to rely upon historical data set, where some important outcome is known. So given a data set with a target variable and input variables, a modeling algorithm automatically generates a model, establishes a relationship between the target variable and those input variables.
This allows us to make predictions, and that's supervised machine learning.
Note: This course is software agnostic. The emphasis is on strategy and planning. Examples, calculations, and software results shown are for training purposes only.
- Evaluating the proper amount of data
- Assessing data quality and quantity
- Seasonality and time alignment
- Data preparation challenges
- Data modeling challenges
- Scoring machine-learning models
- Deploying models and adjusting data prep and scoring
- Monitoring and maintenance