Machine Learning - Introduction

Machine Learning - Introduction

After my latest blog on "Why Robots are the way of the future" I realized I have not really written about one of my greater passions...Machine Learning/Artificial Intelligence.

Machine Learning and Artificial Intelligence are not quite the same thing. Usually Machine Learning falls under the category of Computational Science, Mathematics, Computer Science and Statistics. Artificial Intelligence usually belongs to the Computer Science, Cognitive Science, Psychology and perhaps Mathematical Philosophy departments.

Machine Learning is generally the field of study related to the question: How can we teach a machine to perform a certain task as well as it can be performed?

However, Artificial Intelligence is generally an answer to the question: How can we simulate (or reproduce) the cognitive functions of humans/animals/intelligent beings? A lot of Artificial Intelligence seeks to model the human mind, which doesn't always do things optimally.

Both fields of study are important for developing algorithms and systems that are able to interact, aid, and enhance our lives. For instance, Machine Learning was used to create software to control a refinery. Up until the 1980's most refineries and their myriad pipes, valves, and gauges were controlled by humans. Any potential catastrophes had to be averted by alert workers. However, ML systems were developed that could optimally monitor and control refineries...even better than humans can! Furthermore, as conditions in the refinery change over time the software adapts and retrains itself without having to be reprogrammed over and over.

Interestingly, many of the Machine Learning algorithms are fairly simple in their approach. For instance, a basic classification algorithm is as follows:

Given a set of descriptions x and their associated objects/predictions y which are part of a set of classification categories {y_1, y_2, ... , y_n},

Build a model (a brain)

- For each unique classification y:

-take mean(x) for all x associated with the classification categories y_n

Make Predictions

- take an unclassified description x and evaluate the distances to each of the category means

- the unclassified description x will be classified according to the closest mean

Enhance the model with new examples

-given additional information x and y, recalculate the mean(x)'s with the new information...the model is now enhanced.

For example: Bobby stands by the road and for every vehicle that passes by quickly measures the length of the vehicle. Bobby also writes down whether the vehicle was an "18-wheeler" or "other". Using the above algorithm he calculates the mean "18-wheeler" length to be 20.4 feet, and the mean "other" vehicle length to be 10.9 feet. A blind girl named Jane comes along and says that she has just measured a vehicle that is 16.9 feet long, but can't tell if it's an "18-wheeler" or not. Bobby says that since the unobserved vehicle is closer in length to an "18-wheeler" it must be one!

This may not seem like a very intelligent algorithm (and it's not) but it does demonstrate one key feature of an intelligent algorithm, the ability to form it's internal computing algorithm via external data.

In the next few articles on Machine Learning I'd like to discuss some other more intelligent algorithms such as Support Vector Machines, Neural Networks and Random Forests...all three of which are some of the coolest and most effective AI/ML algorithms. Until then, see if you can think up your own intelligent algorithm...or use for an algorithm. I'd love to hear about it!

Robots are the Way of the Future

Robots are the Way of the Future

I have this really cool robot called Roomba made by iRobot. It vacuums the floor for Malia and I. All we have to do is push the clean button and away he goes zipping around the carpet in a semi chaotic pattern until it's all clean. And it works too!

This is a great development for 2 reasons

1.) It's just cool.

2.) It really does save us time.

Statements like Robots are the way of the future may seem kind of cheesy and cliche. However, I would like to say that there are plenty of cool robots out there. And as Artificial intelligence and Machine Learning develop, robotics will only become more and more useful.

A few months ago on PBS's Nova a program about the Great Robot Race , sponsored by DARPA, was aired in which many vehicles competed to be the first completely machine driven vehicles to make a trek across a desert in California/Nevada. The winner was a small SUV driven without the help of GPS (which is amazing!). It operated on stereo vision, laser vision and a few other sensors, all controlled by a computer with Machine Learning software. Essentially the researchers who created the small intelligent SUV taught the car to drive in the dessert, and it succeeded on the gruelling 150 mile course with an average speed of more than 25 mph!

The significant thing about robots like the winner of the DARPA challenge and Roomba, is that they are actually learning the task set out for them. By learning I mean that the robot's own internal algorithms are being modified as the robot carries out it's task in response to poor or good performance. If you think about it, we learn this way most of the time. In fact, our whole public education system falls under the category that these robot's education falls under: Supervised Learning. (Well ideally at least, a lot of education falls under the category of the 'Expert System', where the teacher poses as an expert and relays his wisdom and knowledge to the student via memorization.)

Supervised Learning is the process where a student is presented a set of examples (descriptions and their corresponding object), and asked to perform a similar identification task on some previously unseen examples. The performance on the unseen examples is evaluated by a supervisor, and the feedback given to the student. The student takes the feedback and reshapes her internal processing algorithms. The process repeats until the supervisor is satisfied with the student's performance.

Robots can even become so good at learning, that they can interface with our brain's functionality! In this story, a man has a prosthetic arm attached to his body and nervous system. Over time, he and the robotic arm are learning to communicate with each other...now that's cool! Star wars is no longer the fantasy world.

I love Machine Learning and Artificial Intelligence, not only because we can use it to make prosthetic arms and automatic carpet cleaners, but because it teaches us something about ourselves. When we explore the world of learning we also question how it is that we know things. We question what knowledge is, how we obtain it, and ultimately what reality is. And at the end of those questions we will always have more, but that's the point of learning!