Artificial intelligence prerequisites?

[icecubebeast]

What are some prerequisites for problem solving, planning and machine learning in artificial intelligence? I was always fascinated by the topic of machine learning until I decided to teach myself how to do it. When I came across a course on machine learning, I was shocked and confused because it had many mathematical topics that I never learned yet.

I've already learned single and multivariable calculus and learned generics in Java. Now, please do not say, "look at the course descriptions of the universities" because when I searched for the prerequisites, the courses never told me what the prerequisites were. I also don't have the patience to go on a search for more than an hour.

 

[axmls]

The listed prerequisites for the undergraduate level artificial intelligence course on MIT's OCW are a knowledge of the Scheme programming language and multivariable calculus. I'm sure another programming language could substitute in another course.

 

[Greg Bernhardt]

Statistics and Probability along with Electrical Engineering

 

[icecubebeast]

The listed prerequisites for the undergraduate level artificial intelligence course on MIT's OCW are a knowledge of the Scheme programming language and multivariable calculus. I'm sure another programming language could substitute in another course.

I checked MIT Opencourseware but it doesn't list the prerequisites. I don't use scheme programming language; I use java. What are some prerequisites for problem solving, planning and machine learning in artificial intelligence?

 

[axmls]

See here: http://ocw.mit.edu/courses/electric...artificial-intelligence-spring-2005/syllabus/

 

[MarneMath]

You're missing the point. I don't use Scheme or Common Lisp for that matter, but I can read Scheme code and write a python code! I recommend moving pass java. Java is a good for some things, but for functional uses, your code is going to be much longer than a perl or python script.

As for what you need for AI. Depends how you define "need". If you are taking the course at a university then the needs can be stripped down. However, if you are self-studying, then before you tackle AI on your own here's a short list of courses that are helpful:

Linear Algebra (At least up to positive definitiness)
Multivariate Calculus
Probability (to some degree simple Statistics such as hypothesis testing)
Numerical Analysis (ie convergence rates, solutions to matrix and simple optimization)
Optimization (convex and nonlinear)

This book below is a good starter that doesn't require much knowledge:
https://www.amazon.com/dp/0136042597/?tag=pfamazon01-20

Now keep in mind the mini-list above is mostly there to provide you a mathematical foundation for AI. However, like most things in life it's very possible to do work within the field without a full expertise on all of this! Especially, if you are interested in learning how to implement some common machine learning algorithms. For example, making a spam filter using a naive bayes would be a good project.

 

[icecubebeast]

To clear it all up, this is what I want to learn and what I want to do with it:
-Machine learning: AI algorithm that "learns" the cause and effect of an event and abstract information about thin
-

@MarneMath Does the book have a good job (detailed and simple language) telling on how to make machine learning algorithms (like what I said above)? I didn't buy the book yet and I don't know the content of it.

 

[MarneMath]

Detailed and simple is relative term. What's obvious and simple for me is probably not obvious and simple for you. That's simply dictated by the fact that I have a broader background than you. Nevertheless, if your interest is simply to create known algorithms then that's trivial and you can essentially google the pseudocode. If your interest lies in understanding search algorithms and reduction problems then this book is pretty solid.

Now you mention machine learning. Machine learning isn't simply AI algorithm. Machine learning is an algorithm that learns from the data it has. It ties to AI, but it isn't equivalent to AI. Below you'll find a few books to cover major topics from ML. I have these books in my bookshelf, but with that said I don't necessary think you're at the level needed for these books, but looking at the content should advise you as to what you need to focus on. (Ie the list I made)
https://www.amazon.com/dp/0387848576/?tag=pfamazon01-20
https://www.amazon.com/dp/026201825X/?tag=pfamazon01-20
https://www.amazon.com/dp/0070428077/?tag=pfamazon01-20
https://www.amazon.com/dp/0521642981/?tag=pfamazon01-20

 

[icecubebeast]

Detailed and simple is relative term. What's obvious and simple for me is probably not obvious and simple for you. That's simply dictated by the fact that I have a broader background than you. Nevertheless, if your interest is simply to create known algorithms then that's trivial and you can essentially google the pseudocode. If your interest lies in understanding search algorithms and reduction problems then this book is pretty solid.

Now you mention machine learning. Machine learning isn't simply AI algorithm. Machine learning is an algorithm that learns from the data it has. It ties to AI, but it isn't equivalent to AI. Below you'll find a few books to cover major topics from ML. I have these books in my bookshelf, but with that said I don't necessary think you're at the level needed for these books, but looking at the content should advise you as to what you need to focus on. (Ie the list I made)
https://www.amazon.com/dp/0387848576/?tag=pfamazon01-20
https://www.amazon.com/dp/026201825X/?tag=pfamazon01-20
https://www.amazon.com/dp/0070428077/?tag=pfamazon01-20
https://www.amazon.com/dp/0521642981/?tag=pfamazon01-20

Sorry, I forgot to mention in the previous post (computer was crap and internet was buggy).

So another thing I want to learn is to make AI programs that can understand how certain actions cause certain events and how to use it. By "use it" I mean that they can build algorithms and self improve (by "improve" I mean more efficient and more capable code) the code in the program that makes it (Java).

 

[icecubebeast]

@MarneMath
What prerequisites do you need for language interpretation and understanding?

 

[donpacino]

Look up neural networks. It is a base course for machine learning.

Also look into fuzzy logic

 

[icecubebeast]

Look up neural networks. It is a base course for machine learning.

Also look into fuzzy logic

Can you give a list of books on amazon, google play, itunes, etc.?

 

[donpacino]

Can you give a list of books on amazon, google play, itunes, etc.?

http://lmgtfy.com/?q=neural+network+books

 

[icecubebeast]

http://lmgtfy.com/?q=neural+network+books

Can you please give me a LIST of books on neural networks that you recommend? Your reply does not answer my question.

 

[donpacino]

I don't know enough about the differences between the books to give you a recommendation. all i would do is google neural network books and give you the highest rated ones.