Obviously, people define AI and machine learning in many different ways. In 2019, it is still unclear what AI is capable of and what the exact definition is.
In our last post on creative data predictions, we defined AI like so:
Artificial Intelligence: A non-human system that shows human-like intelligence. AI is an umbrella term, which includes machine learning and other techniques. Examples include playing the computer on a video game or talking to Siri.
We also noted, “ML is AI but AI is not ML.” Take a minute. Reread. Okay, let’s go.
We defined machine learning as:
Machine Learning: A method to teach an AI system to perform a task based on the data given to them. Basically, you are "training" a "model" using ML techniques to make predictions. (Often this model is also called an ML model). Examples include Waze giving you the best route home based on user traffic data.
We even made this nifty graphic for you to save and refer to:
And we jotted down some other definitions to easily digest as well.
What are Machine Learning Predictive Models?
The terms “Machine Learning” and “Predictive Modeling” can be used almost interchangeably. Data analysts use Machine Learning to build predictive models to show the most likely possible outcome.
What is Predictive Analytics?
Predictive analytics is the analysis of data to see patterns and relationships between variables to identify future outcomes.
But it’s a new post, it’s a new day. Let’s go deeper into what AI and ML are and why you need them in your 2019 workflow to be creative with business solutions.
AI is, Like, the Future?
Yes, you could say that. And if you have been caught explaining AI like this, you’re not alone. We all vaguely explained how AI is the future at one point or another. It’s hard to explain how AI will affect our lives. Only one critique of this statement: AI is the future, but it’s also the present. Repeat that in the next conversation with your friends about AI and you’ll blow some minds.
There Are Many Ways to Use AI
Types of AI include:
- Big Data
- Data Mining
- ANI (Artificial Narrow Intelligence)
- Computer Vision
- NLP (Natural Language Processing)
- Machine Learning
Let’s Focus on Machine Learning's Relationship With AI.
We’re only going to focus on ML and it’s subsets for this post. However, we will have a post on how NLP is coming to the Obviously AI platform in the future. (If you haven’t already, sign up for our newsletter below this post to stay in the know.)
Like I said before, ML is AI, but AI is not ML. AI is an umbrella term for many facets of tech.
Machine learning appears all around us in our day-to-day lives. Spotify and Netflix use big data to build predictive models to present you with suggested content and improve UX. Social media apps present ads based on your browser history. Google Maps guesses where you’re going before you type in an address based on your past data.
You can apply ML to perform tasks in many different ways, but here are some ways it’s used in business:
Deep Learning: Deep learning is where a system takes thousands of data points and teaches itself patterns from the data. This can include images, video, music, text or other samples and classify them as a human mind would.
Predictive Modeling: Using data to build predictive models to predict the most likely outcome. Predictive models identify patterns and relationships between the variables. This is what we do—except you don’t need to know how to program and use algorithms. You can just ask our platform a question in plain English and it’ll apply the best algorithm to get the answer.
Customer Relationship Management (CRM): CRMs compile data from different aspects of a business (websites, social media, content, emails, etc.) to help the user better understand their audience better and improve sales and customer retention.
Neural Networks: A pattern-recognizing algorithm can create an artificial neural network allowing a computer to learn from and figure out sensory data to cluster and classify it. You’ll see this mostly in autonomous cars who learn from their environment, reading street signs, pedestrians, etc., and apply it to their future decisions.
There are tons of others as well that we won’t get into in this post. Things like search engine optimization, biomedical informatics, data mining, email filtering and such will have to wait.
There Are Two Main Methods to Machine Learning
Supervised Learning is a method that has an input and output variable. Problems using this method can be grouped into regression or classification.
Unsupervised Learning is a method that has an input, but no output variable. This method is mainly used to learn more about the data and present interesting relationships about the variables. This method can be grouped into association or clustering.
Wait, What About Predictive Analytics?
I mentioned this in my last post, but it’s worth repeating. “Predictive modeling” and “machine learning” can be almost used interchangeably. While they have minor differences, they both encompass using data to perform a task. Predictive analytics is another term that’s related and you might see used in the same way as machine learning and predictive modeling.
Predictive analytics is the analysis of data to see patterns and relationships between variables.
Predictive Analytics Has a Process.
- Collect Data
Also in our last post, we go into greater detail of what predictive analytics can do and how it’s used in business.
For the Next Post, We Will Discuss How to Add ML and Predictive Analytics to Your Workflow.
And I’m employing a special guest to help narrate the next post. If you have a hunch about who it will be, reach out to us on Twitter.
Hint: He is a fellow Berkeley resident and recently released an album (with others) using ML, creating the perfect segue into how to incorporate ML creatively into your workflow.
We will publish the post next Monday. Peace. ✌️