Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 6 years ago.
Improve this question
I have a homework in artificial intelligence.
I need to make a robot go from room A to room B and there are obstacles between the rooms.
The professor asked me to use STRIPS (Stanford Research Institute Problem Solver) but I can not understand how STRIPS works.
Can someone give me a good explanation and examples about what is strips and how it work?
Thank you.
[Please note, this is based on what I half-remember from nearly a year ago]
These days, I would expect that when the Prof. says STRIPS, they would be talking about the problem coding 'language', rather than the planner - check for example the Wikipedia page: STRIPS. I would imagine that your Prof. likely has a particular solver (and quite possibly algorithm too) in mind, and is wanting you to encode the domain and specific problem, to run on the solver. Without knowing more details of the assignment, I can't be sure what you need. If you're looking for a planner, as I understand it, Fast Downward is quite popular among researchers currently. The website has some instructions on how to use it, and IIRC it comes with a bunch of domains and problems for those domains. I would thoroughly recommend looking at those, they're pretty much what I learnt with. I also just found this and this.
STRIPS is essentially a way of encoding information about the nature of the problem you want the computer to find a solution to. Typically you encode a domain, which provides information about the problem overall, such as what objects may be involved, what states they can be in, and what actions can be taken. Then, you also encode a particular problem, which (generally) specifies the starting state of the problem, and what the goal state should look like. Both those files are fed into a solver, which takes them and then finds a solution to the problem. Note that this won't always necessarily be an optimal solution - that depends on what algorithm you use, and how you have told the solver what should be optimised (which I think you can generally do in the problem, though I can't remember for sure now).
I suggest you have a look at those links, and see what you can find out. That should hopefully give you a better idea of what gaps you need filled in your knowledge, and then you can narrow in on exact specifics. If this is a taught course assignment, then I would expect that surely the Prof. would have gone over some of this in lectures (do you have lecture slides available?), or at least pointed everyone towards a recommended planner and material to read up on. If you're still struggling, your best bet is to go back and see the Prof. in office hours.
Related
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 2 years ago.
Improve this question
PostgreSQL makes use of intraoperation parallelism and that is of interest to me (for my undergad final year research project). I would like to know how operations like selection, projection, join, etc are parallelized, but when I tried to look at the source code, I got extremely overwhelmed. Is there a high-level PostgreSQL "map"?
I tried looking for books that discuss and explore the algorithms and implementations used in PostgreSQL, but unfortunately didn't find any. Though feel free to refer me to such a book if you know about one.
If the only option I have is to dig into the source code, how long would it take me to find the information I want? And if any of you have gone through the source code, what advise would you give to me?
The nice thing about open source is that there is no clear border between the source code and the documentation, since both are public. As soon as you get deeper into the implementation details, you will start reading the code. Fortunately the PostgreSQL code is well written and quite readable.
The first stop on your way into the source are the README files. These describe implementation principles, algorithms and code rules at a higher level. In your case, you should start with src/backend/access/transam/README.parallel.
Another good approach it to read the patches that introduced the feature, like 924bcf4f16d, 7aea8e4f2daa, d1b7c1ffe72, f0661c4e8c44 and 80558c1f5aa1. That introduces you to the places in the code that are concerned with parallel query and gives you an idea how it all works.
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 8 years ago.
Improve this question
Suppose I've been recruited to a university study testing a new AI, and we're going to do a traditional Turing test. I will sit at a computer and interact via text-based chat with something that is either a computer or another volunteer, and then have to guess which it is.
Suppose I've been given a particularly nice incentive to get it right. How can I improve my odds? What sort of things should I talk about?
Should I try to teach the chatter a simple new concept I just made up, and then quiz them about it?
Should I try to make jokes and innuendo and see if they "get it"?
Should I ask philosophical questions?
Should I try practical questions that humans tend to "irrationally" answer in a particular way?
Try some l33t speak, that might do it. Then again I dont even understand some of it at times. Maybe I'm a computer. Maybe you're a computer
Realize that the state of the art in chatbots is still abysmal. The biggest mistake you can make is to try to do something bizarre, in which case both a human and an AI might reasonably respond in a bizarre manner. Instead, just engage in a normal conversation and resist any significant deviations proposed by your correspondent. That should easily let you identify anything operating at the current published state of the art.
If you encounter something well beyond state of the art, I think the next threshold would be a directed inquiry into a aesthetic decision: "What's your favorite [movie|book|artist|meal]? Why? So would you say you never like...? Are there aspects of that you dislike...?" etc.
Since computer programs might be very logical and high performing, but have no experience of being a human being, I would suggest questions of type trivial everyday events that wouldn't get adequate answers on Google, likely. As questions about cleaning, small details on outfits, boring and irritating situations ...
And it might be a good idea to communicate in a spoken careless way with lots of words and sentences with clear and controversial statements of the above type only here and there.
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 9 years ago.
Improve this question
I am searching for a month for blog posts about "how C program works". Most of them go like
Compilers do these things
Linkers do these things
Program is put into memory; and stacks, heaps, etc.
I thought I would first read about how compiler works to understand the flow of the program into the machine. Dragon Book seems to be universally preferred. But truthfully, it's too intensive. I am not good enough now to go through it all.
So I began to read about hardware. But There too, they explain about buses, I/O signals, structure of memory, writing cache-friendly code etc. But with no proper examples.
But still I could not find myself satisfied or being able to completely visualize the process.
2 hours ago I decided to ask this question.(Since I am scared of It might be not useful to SO community, or off-topic question or other down-votable categories) and I did not find out any post relating to this exactly. There was one about "how compiler does the compilation", but the answers showed that it's too broad a question.
My question is this:
I would like to know how, in depth, a C program works. If you cannot tell me explicitly, please redirect me to a book or another post on another website that can give me the answer to this.
I am here until I get a response. If you have any suggestions regarding this post, tell me. And It is not my first language, so please take all my sentences as being soft and polite.
Thanks.
UPDATE:
Along with the accepted answer, there are some very nice links as well as suggestions which give partial answers or the way to proceed further to understanding what I am trying to understand.
The best answer to this question by far comes from the book "The Elements of Computing Systems," by Noam Nisan and Shimon Schocken. This book starts from the simplest possible electronic components, assembles them into a working processor, invents a simple assembly language for it, writes an assembler for that, and ultimately shows you how high-level languages can be compiled onto it. Reading the book, and working all the examples (which use a simulator for the hardware, so no workshop required!), will forever change the way you look at computers; he will literally understand everything from the lowest to the highest levels, and see how they work together. See the book's website for more info.
It's too broad a question (as you have observed).
If you really want to understand from bottom up - buy an OLD computer from the 80's off ebay. Sinclair Spectrum/BBC it really doesn't matter but make sure you get plenty of books and manuals that go with it.
You will learn plenty because these machines were well documented and what wasn't documented was discovered and then documented :)
They are also sooooo much simpler than a modern quad core multi-gigabyte memory job. It will all fit inside your head easily
or for a modern start, Arduino or Raspberry pie maybe.
Closed. This question is off-topic. It is not currently accepting answers.
Want to improve this question? Update the question so it's on-topic for Stack Overflow.
Closed 9 years ago.
Improve this question
How would you engineer an impossible bug? Is it even possible?
Take a simple example: Bob needs to add commenting functionality to his popular internet blog. It gets millions of visitors a month and already has thousands of posts. How would you implement the feature if he hired you?
You would probably add a form at the end of the page which is responsible for rendering posts, right? Its almost common sense. (Forget about security for the sake of simplicity)
However, that solution assumes way too much about how Bob's blog works.
Let's say that the person who made Bob's blog hated Bob. He wanted to make a system where no new features could be implemented without completely rewriting it from scratch.
How would you create a system like that if you were the one that hated Bob? You aren't limited to a specific language, framework, operating system, server software, architecture, or anything. All you need to do is provide working blog pages to clients that connect to your server.
Also assume Bob will do whatever you tell him to. If he has to send in his new hand-written blog posts by carrier pigeon to an offshore processing center, then he will do it.
And before anyone asks, NO, I am NOT asking this question because I want to screw over a client. I think of this like an exercise. We're used to engineering code that isn't supposed to break, but how would we do engineering code that isn't supposed to work? Just like bugs are inevitable, are solutions equally inevitable? Is it just as hard to create inverted code like this as it is to write regular code? If we can create code that can't be fixed, does it mean we can create code that can't be broken?
Because the blog platform itself is fairly expendable, I would think that the only sort of bug that irreversibly cripples Bob's web-presence would have to be a bug targeted at either the content, or at Bob himself. Some examples of this would be placing the only terminal that allows Bob to post in a cage full of lions, or convincing the internet to DDoS every blog written on the same topic as Bob's blog.
That said, as long as Bob is a citizen of the free world where free internet access in public cafes and libraries abounds, I don't think it will ever be possible to introduce a bug so terrible that Bob's blog cannot be published.
Let's say that the person who made Bob's blog hated Bob. He wanted to make a system where no new features could be implemented without completely rewriting it from scratch.
How would you create a system like that if you were the one that hated Bob?
I would create the blogging system in Malbolge.
Closed. This question is opinion-based. It is not currently accepting answers.
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed 8 years ago.
Improve this question
When you are watching for news of particular Wikipedia article via its RSS channel,
its annoying without filtering the information, because most of the edits is spam,
vandalism, minor edits etc.
My approach is to create filters. I decided to remove all edits that don't contain a nickname of the contributor but are identified only by the IP address of the contributor, because most of such edits is spam (though there are some good contributions). This was easy to do with regular expressions.
I also removed edits that contained vulgarisms and other typical spam keywords.
Do you know some better approach utilizing algorithms or heuristics with regular expressions, AI, text-processing techniques etc.? The approach should be able to detect bad posts (minor edits or vandalisms) and should be able to incrementally learn what is good/bad contribution and update its database.
thank you
There are many different approaches you can take here, but traditionally spam filters with incremental learning have been implemented using Naive bayesian classifiers. Personally, I prefer the even easier to implement Winnow2 algorithm (details can be found in this paper).
First you need to extract features from the text you want to classify. Unfortunately the Wikipedia RSS feeds don't seem to be particularly machine readable, so you probably need to do some preprocessing. Alternatively you could directly use the Mediawiki API or see if one of the bot frameworks linked at the bottom of this page is of help to you.
Ideally you would end up with a list of words that were added, words that were removed, various statistics you can compute from that, and the metadata of the edit. I imagine the list of features would look something like this:
editComment: wordA (wordA appears in edit comment)
-wordB (wordB removed from article)
+wordC (wordC added to article)
numWordsAdded: 17
numWordsRemoved: 22
editIsMinor: Yes
editByAnIP: No
editorUsername: Foo
etc.
Anything you think might be helpful in distinguishing good from bad edits.
Once you have extracted your features, it is fairly simple to use them to train the Winnow/Bayesian classifier.