Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

How can one manage thousands of IF…THEN…ELSE rules?

programmers.stackexchange.com

Post reply on HN
Print view View original

21–30 of 86 posts

Re: How can one manage thousands of IF…THEN…ELSE rules?

#21

Post by instakill »

The question is an interesting one but it's futile in its context. You can't build a prediction engine for cattle movement. Adding factors like weather, food etc. seems well and dandy but there's hundreds if not thousands of other factors that are definitely going to be missing and that will render the prediction events pointless.

Never mind the fact that you can throw in black swan type of events into the mix or just unknown unknowns you could never think of (maybe a cow has a frog phobia and panics when it sees the frog and starts a stampede or whatever it is that cattle do).

Re: How can one manage thousands of IF…THEN…ELSE rules?

#22

Post by _delirium »

While I'm not sure thousands of if/then/else rules are actually the best knowledge representation for his domain (it sounds like he really wants to learn some model from data, in which case throwing some off-the-shelf machine learning may be a better fit), the area of "expert systems" may be relevant if they really are rules extracted directly from some source that need to be managed, such as a human domain expert.

Expert systems were particularly big in AI in the '80s, and considerably less hot now, but still widely used in industry. There are a lot of issues that come up, most of which aren't really purely technical, such as: 1) how you extract knowledge from people who might know it; and 2) how you validate that the knowledge you've extracted is actually what they know, for example by validating that it produces the same decisions that they would make; and 3) how you allow updates/revisions to the knowledge base over time. Once you have the rules and some reason to believe that they're any good, actually managing and applying them can be done through one of several rule engines, such as Drools or Jess.

The keyword "knowledge engineering" may also turn up relevant info.

Re: How can one manage thousands of IF…THEN…ELSE rules?

#23

Post by femngi »

SeanLuke wrote: post #13
Dn_Ab wrote: post #4

Such a great question. Want to predict cow movement while taking into consideration sudden events, food and sun, using thousands of if then rules? Sounds like the perfect problem for one of the following tree learners that not only manage themselves but build the model for you, in order of suitability to the problem* : Genetic Programming ADTree Random Forest Decision Tree Most of the work would be in figuring out th…

All machine learning techniques build models for you. ADTrees, Random Forests, and Decision trees are all classifiers, so they serve this guy's needs. Genetic Programming is not a machine learning technique: it is an optimization method for a very specific task and is quite unsuited for his purposes.

Can you explain you think Genetic Programming is unsuitable? I honestly can't see why you think it wouldn't work.

Re: How can one manage thousands of IF…THEN…ELSE rules?

#24

Post by StavrosK »

femngi wrote: post #23
SeanLuke wrote: post #13

Earlier quoted context omitted.

All machine learning techniques build models for you. ADTrees, Random Forests, and Decision trees are all classifiers, so they serve this guy's needs. Genetic Programming is not a machine learning technique: it is an optimization method for a very specific task and is quite unsuited for his purposes.

Can you explain you think Genetic Programming is unsuitable? I honestly can't see why you think it wouldn't work.

As SeanLuke said, genetic programming is an optimization method. I don't see a function being optimized here, so what would you run it on?

Re: How can one manage thousands of IF…THEN…ELSE rules?

#25

Post by jetti »

ynniv wrote: post #12
sliverstorm wrote: post #9

Earlier quoted context omitted.

What, are you insinuating that asking about great big if-then-else trees means you'd never hire that person? Because, what, you've never been a complete novice programmer before? Considering the question-asker has a specific problem to solve, I would even hazard they probably aren't a programmer by trade.

Considering the question-asker has a specific problem to solve, I would even hazard they probably aren't a programmer by trade. So why would someone hire them to be a programmer? And yet, unlike other skilled professions, rank amateurs often present themselves as competent programmers. It would be valuable to include evidence to the contrary with résumés. you've never been a complete novice programmer before? I certa…

So you rather hire somebody who has the same problem but has too much pride to ask somebody for help? Seriously, if all the questions everybody asked ever were on their resume, nobody would ever be able to get a job. At least this individual recognized that there was something wrong and sought out how to fix the problem from people with more experience then them.

Re: How can one manage thousands of IF…THEN…ELSE rules?

#26

Post by ck2 »

Is it just me? It's not a good question, it shows lack of core coding experience.

For example a video game consists of millions of possibilities/results in movement based on environment - it's not approached as a collection of thousands of if/then/else statements.

In fact the movement of cows seems very much like video game coding, it needs a "cow engine".

Re: How can one manage thousands of IF…THEN…ELSE rules?

#27

Post by dave_sullivan »

This really sounds like a machine learning problem--why hand engineer all these rules yourself (which can be error prone in any case) when a program could do it for you? If you can figure out how to get a large data set together and come up with a good general representation of the data, you're half way there.

I'd consider taking a look at recurrent neural networks if you're looking at your data as a time series or if its more of a static problem that doesn't take into account the last N things that came before, you might consider tree based methods or even DBNs if you can get a lot of data together (say 50000-100000 samples)

Potential pitfall: if you're using a NN based approach, you will judge results based on test performance, but you won't get as much insight into the "rules" your network has learned.

Re: How can one manage thousands of IF…THEN…ELSE rules?

#28

Post by ynniv »

jimbobimbo wrote: post #10
sliverstorm wrote: post #9

Earlier quoted context omitted.

What, are you insinuating that asking about great big if-then-else trees means you'd never hire that person? Because, what, you've never been a complete novice programmer before? Considering the question-asker has a specific problem to solve, I would even hazard they probably aren't a programmer by trade.

I'd say the fact that person is asking this question is a very good sign - they have a gut feeling that something's off with IF-THEN spaghetti.

... a very good sign that they're still too green.

When did HN become so touchy-freely? It's important that people start learning somewhere. If you show any interest in programming, I highly recommend learning more. However, there comes a time when you must have a certain amount of experience. This is a profession where important things are made. Even people who should know better end up building systems that leak our sensitive personal information, or let others impersonate us.

I didn't say that such a person shouldn't be hired, but I certainly implied that the original question is a better indicator of their capability than the always positive keyword soup and "years experience" that many people slather on their résumé.

If "avoiding IF-THEN spaghetti" is all you require from contributors, fine. Let's not pretend that this is better than an early high school effort.

Re: How can one manage thousands of IF…THEN…ELSE rules?

#29

Post by Evbn »

instakill wrote: post #21

The question is an interesting one but it's futile in its context. You can't build a prediction engine for cattle movement. Adding factors like weather, food etc. seems well and dandy but there's hundreds if not thousands of other factors that are definitely going to be missing and that will render the prediction events pointless. Never mind the fact that you can throw in black swan type of events into the mix or jus…

Yeah, it is as silly as trying to predict the weather.

Re: How can one manage thousands of IF…THEN…ELSE rules?

#30

Post by ynniv »

jetti wrote: post #25
ynniv wrote: post #12

Earlier quoted context omitted.

Considering the question-asker has a specific problem to solve, I would even hazard they probably aren't a programmer by trade. So why would someone hire them to be a programmer? And yet, unlike other skilled professions, rank amateurs often present themselves as competent programmers. It would be valuable to include evidence to the contrary with résumés. you've never been a complete novice programmer before? I certa…

So you rather hire somebody who has the same problem but has too much pride to ask somebody for help? Seriously, if all the questions everybody asked ever were on their resume, nobody would ever be able to get a job. At least this individual recognized that there was something wrong and sought out how to fix the problem from people with more experience then them.

Meh, put a date on it. If you asked this even a couple of years ago, fine. If you asked this last month, I have a pretty good feel for where you're at. You should also learn from your questions, which would make a great interview conversation.

I've littered the net with my share of dumb questions, and I have no regrets.

Post reply on HN