Assembly Line Scheduling - Data structures - data-structures

I am trying hard to understand how the table is computed in this 'Assembly line scheduling' problem in Chapter 15 'Dynamic Programming' on Intorduction to Algorithms book by Cormen.
Can anyone give me a hint on what are the two tables about & how they get calculated ? Been searching on google for past 2 hours.

The 1st table contains the fastest way to pass station Si,j:
f1[j] is the fastest way to pass station j of assembly line 1
f2[j] is the fastest way to pass station j of assembly line 2
For example, f2(3)=22, since 2+7+2+5+6=22, and this is the fastest way to pass station 3 of assembly line 2.
In the 2nd table, li(j) shows the assembly line number (1 or 2) that is used in step j-1 as part of the fastest way to reach li(j).
For example, l1(2)=1 since the fastest way to reach station 2 in assembly line 1 is through station 1 on assembly line 1. (2+7 < 4+8+2)
l2(2)=1 since the fastest way to reach station 2 in assembly line 2 is through station 1 on assembly line 1. (2+7+2 < 4+8).

f(2) = 22 because it is the shortest way (time) to pass station 2,3 :) You have to look all ways and find shortest f(1)j is station 1,j I know because I took this lesson too :)

Sorry for digging up a 4-year-old question, but it's the first result on google.
The answer provided by Lior is incorrect. f1(j) is NOT for passing stations starting at assembly line 1. If so, why is f1(2) = 18? when the optimal path is 2+7+2+5= 16.
Also, for f2(3) = 22, 4+8+5+1+3 does NOT equal 22. It's 21.
fi(j) is actually the function of the fastest way to get to jth station on ith line (as answered by Kubra).
f2(3) = 22 because 2+7+2+5+6. That's the most efficient route to get to that specific station.
I hope my answer will save people's time, as i spent an hour on double-, triple-checking if i made a mistake understanding the problem and answers.
Thanks.

Related

Information Retrieval - Adjancey Matrix Graph Sketch, Teleportation Probability, Calculate PageRank

I am doing a few thing on Information Retrieval and have an exam coming up and I am absolutely clueless. First of, could anyone recommend me the shortest and best description possible for what PageRank actually is in Information Retrieval? Maybe even a good short video or your own description. I know Google use to, or did use it.
I know there are a lot of questions here but I could use as MUCH help as possible in a short length of time.
So my first question (taken from past papers, and making my own examples):
I am wanting to take a table such as:
A B C
A 0 1 0
B 1 0 1
C 0 0 0
And create a graph. I believe this is correct but unsure (I could use a "yes that is correct" or a "no":
And if I was given a graph such as:
The table would be:
A B C
A 0 1 0
B 0 0 1
C 0 0 0
Is that correct? If not, could I please get help and get it described somewhere? The lecture I am reading is not great at explaining and my lecturer isn't great at helping either.
Next I will probably be asked to use Teleportation Probability on the first table. This I desperately need help in. If the probability(the special a symbol)=1/2, does this mean multiply everything, including the 0's in the table such as 0x1/2? also 1x1/2? This is for the matrix of transition probabilities.
Next would be, how can I calculate PageRank from the above matrix. Using matrix multiplication. In words or in Pseudocode.
Another question I want to know is, will a user's page rank on twitter increase if they follow another user? I was assuming this would be a no because they are not following the user back?
Does a user's pagerank depend on how frequently you find said user if you start at a random user and click on another random persona and such till you find them? I assume this one is definitely not true. Because they might not be following said user.
I know this is a lot to ask. Does anyone have tutorials I can follow for either that are not complicated and I can look at and get it mastered today?
Thanks I really appreciate all your help. I know not one person can answer them all but can help provide assistance for some.
here's my stab at answering your questions:
good learning resource:
http://en.wikipedia.org/wiki/PageRank#Simplified_algorithm (no doubt you've see it already, but it's a pretty good one). Start there, understand the algorithm first, then do the implementation.
this might be a good simple method to implement?
http://pr.efactory.de/e-pagerank-algorithm.shtml
or this:
http://www.cs.princeton.edu/~chazelle/courses/BIB/pagerank.htm
I'm guessing you can program in Python (common school language), in that case you might be interested in a package for handling graphs which has pagerank calculations: http://networkx.lanl.gov/reference/generated/networkx.algorithms.link_analysis.pagerank_alg.pagerank.html. If you have to write your own pagerank algorithm (very doable), you could use that to check the results.
For the matrix -> graph conversion question: your professor needs to specify how directionality is encoded in the matrix. Does a 1 at B,C specify a link from B to C or from C to B? My guess would be B to C. If that's true, your first graph is wrong there, but the second graph is ok. Directionality is very important in PageRank.
I believe the Teleportation probability is the probability that a random walker executing a new step will jump to a random node in the graph. It's in the wikipedia page under "damping factor". I don't know how it ties into multiplying numbers in your matrix.
For the Twitter question - yes, I think you have it right. Linking to (or presumably following) a second person does nothing directly to the the first person's pagerank, but it likely increases the second person's pagerank. In practice, there could be secondary effects, like the second person noticing that the first person is interesting and following them back.
second to last question - yes, one formulation of the pagerank algorithm is as a random walk along links with the frequency of encountering a node (page) going into the pagerank.
good luck!

Optimal sequence to brute force solve a keypad code lock [duplicate]

This question already has an answer here:
Closed 10 years ago.
Possible Duplicate:
Need help in building efficient exhaustive search algorithm
Imagine that you must open a locked door by inputting the correct 4-digit code on a keypad. After every keypress the lock evaluates the sequence of the last 4 digits inputted, i.e. by entering 123456 you have evaluated 3 codes: 1234, 2345 and 3456.
What is the shortest sequence of keypresses to evaluate all 10^4 different combinations?
Is there a method for traversing the entire space easy enough for a human to follow?
I have pondered this from time to time since a friend of mine had to brute force such a lock, to not having to spend the night outdoors in wintertime.
My feeble attempts at wrapping my head around it
With a code of length L=4 digits and an "alphabet" of digits of size D=10 the length of the optimal sequence cannot be shorter than D^L + L - 1. In simulations of smaller size than [L,D] = [4,10] I have obtained optimal results by semi-randomly searching the space. However I do not know if a solution exists for an arbitrary [L,D] pair and would not be able to remember the solution if I ever had to use it.
Lessons learned so far
When planning to spend the night at a friends house in another town, be sure to not arrive at 1 am if that person is going out to party and won't hear her cell phone.
I think you want a http://en.wikipedia.org/wiki/De_Bruijn_sequence - "a cyclic sequence of a given alphabet A with size k for which every possible subsequence of length n in A appears as a sequence of consecutive characters exactly once."
The link Evgeny provided should answer both of your quests. This answer is a bit offtopic, but you ask for a solution for humans.
In the real world you should probably rely more on Social engineering or heuristics, and after that on mathematics. I give a case on real life:
I went to visit an apartment and I found out that my cellphone was dead. Now way of contacting the person doing the visit. I was about to go back when I saw that the door used a keypad 0 - 9 and A B. I made several assumptions:
The code is 5 digits long. The length is pretty standard depending on the region you are in. I based this assumption on buildings I had access before (legally :D).
The code starts with numbers, then either A or B (based on my own building).
The keypad was not brand new. Conclusion, the numbers used in the code were a bit damaged. I knew with certainty which numbers were not in the code, and three of the four number in the code (given my previous assumptions)
By the amount of keys damaged I assumed the code didn't contain repeated keys (7 were damaged, it was clear A was used, B not used )
At the end I had 3 numbers which were in the code for sure, 2 candidates for the last number and I was sure A was at the end. On key was just slightly damaged compared to the others.
I just had to enumerate permutations starting with the candidate which seemed the more damaged, which give me 4! + 4! = 48 tries. Believe me, at the 5th try the door was opened. If I can give my 2 cents, the old put a key and open the door is still the most reliable method to restrict access to a building.

Generating Bulls and Cows secret word with given n String inputs

I am stuck on this problem for quite a while, it is basically reverse engineering bulls and cow game.
Read more here: http://rosettacode.org/wiki/Bulls_and_cows
I am not able to develop a logic for the problem given below, if you can think of a solving approach please comment the same.
Problem Statement:
Given few clue words(of form ABCD/DBCA etc) and the number of cows and bulls for each word,program
should be able to work out the actual word by evaluating the given clue words and generate the output secret word.
TEST CASES:
Input:
4
DBCC 0 2
CDAB 2 1
CAAD 1 2
CDDA 2 0
Output:
BDAA
The idea is to reduce the space of possible solutions. Before you start, all 4^4 combinations are possible. After you read the first clue [DBCC 0 2 ], you can eliminate a number of possible solutions, in this particular example you can eliminate all states which have a D in the first place, all which have a B in the second place and so on. Just eliminate all possible solutions which do not "fit" the current clue.
Do this with each clue, until only one solution is left. Another interesting problem of course is how to generate good clue patterns.
The way I did it is:
1. Generate all possible words, put them in a list (array)
2. Randomly select one of them (first question)and ask for clues
3. Take the answer (let's say it is 2,1)
4. Start comparing that question with
first, second, ..., to the last word from the list
5. if they give the same clue: count them, plac

Need help to adapt a genetic algorithm to "solve" Traveling Salesman on Ruby

I just downloaded ai4r library http://ai4r.rubyforge.org/ and i am using the genetic algorithm to get a good route from multiple places, just like this:
http://ai4r.rubyforge.org/geneticAlgorithms.html
But i need to be able to set a start city.
Any clue on how to use this on a "fixed" start city?
In general, not looking at the Ruby-specific implementation, you can just remove the start city, and assume that the first edge followed is from your start city to whatever the GA produces. Just make sure that first edge is included in the cost/fitness function.
The solution of a traveling salesmen problem is a round-trip and is therefor independent of a start-city. After you have found your solution you can take any city of the round-trip as your start-city.
EDIT: If you do not need to return to your start-city, you can select the end city, by removing the larger of the two distances that leave your start city. If you remove in your final solution the largest distance in the whole round-trip you get the overall shortest tour that is not a round-trip. This is likely what they did on the web page you linked (Dublin - Moscow looks to be the most expensive direction). However, note that the authors of that page used a wrong location for Vienna and Madrid seems to be off as well.
Another way of when you'll be needing a start-city is when you have an additional time window constraint. This constraint specifies that for each city you need to be there at a certain time, the "depot" as your start-city is called in this case has a time window that covers the whole trip. However the TSPtw is a much more complex problem and often requires advanced genetic operators. You can also model the TSPtw as a CVRPtw (Capacitated Vehicle Routing Problem with Time Windows) if you use just one vehicle. You can try our VRP implementation in HeuristicLab to solve this problem. We have a mailing list if you require further support.
The answer you'll get from the GA to the TSP is a cycle, that means that the first city is the on that you desire. For example, if the answer is [3 4 2 6 1 5], and I want the first city to be 2 then I can "roll" the solution to [2 6 1 5 3 4].
Although, you can reduce your problem by 1 if in your evaluation function you specify the first city. In that case you must take into account that the individuals must be modified to account this. For example, if you set the first city to #2 (problem of 6 cities) and you have an individual that is [1 2 3 4 5] (6 cities minus one). The individual to evaluate is [2 1 3 4 5 6].

What class of algorithms can be used to solve this?

EDIT: Just to make sure someone is not breaking their head on the problem... I am not looking for the best optimal algorithm. Some heuristic that makes sense is fine.
I made a previous attempt at formulating this and realized I did not do a great job at it so I removed that question. I have taken another shot at formulating my problem. Please feel free to provide any constructive criticism that can help me improve this.
Input:
N people
k announcements that I can make
Distance that my voice can be heard (say 5 meters) i.e. I may decide to announce or not depending on the number of people within these 5 meters
Goal:
Maximize the total number of people who have heard my k announcements and (optionally) minimize the time in which I can finish announcing all k announcements
Constraints:
Once a person hears my announcement, he is be removed from the total i.e. if he had heard my first announcement, I do not count him even if he hears my second announcement
I can see the same person as well as the same set of people within my proximity
Example:
Let us consider 10 people numbered from 1 to 10 and the following pattern of arrival:
Time slot 1: 1 (payoff = 1)
Time slot 2: 2 3 4 5 (payoff = 4)
Time slot 3: 5 6 7 8 (payoff = 4 if no announcement was made previously in time slot 2, 3 if an announcement was made in time slot 2)
Time slot 4: 9 10 (payoff = 2)
and I am given 2 announcements to make. Now if I were an oracle, I would choose time slots 2 and time slots 3 because then 7 people would have heard (because 5 already heard my announcement in Time slot 2, I do not consider him anymore). I am looking for an online algorithm that will help me make these decisions on whether or not to make an announcement and if so based on what factors. Does anyone have any ideas on what algorithms can be used to solve this or a simpler version of this problem?
There should be an approach relying upon a max-flow algorithm. In essence, you're trying to push the maximum amount of messages from start->end. Though it would be multidimensional, you could have a super-sink, which connects to each value of t, then have each value of t connect to the people you can reach at this time and then have a super-sink. This way, you simply have to compute a max-flow (with the added constraint of no more than k shouts, which should be solvable with a bit of dynamic programming). It's a terrifically dirty way to solve it, but it should get the job done deterministically and without the use of heuristics.
I don't know that there is really a way to solve this or an algorithm to do it the way you have formulated it.
It seems like basically you are trying to reach the maximum number of people with exactly 2 announcements. But without knowing any information about the groups of people in advance, you can't really make any kind of intelligent decision about whether or not to use your first announcement. Your second one at least has the benefit of knowing when not to be used (i.e. if the group has no new members then you can know its not worth wasting the announcement). But it still has basically the same problem.
The only real way to solve this is to use knowledge about the type of data or the desired outcome to make guesses. If you know that groups average 100 people with a standard deviation of 10, then you could just refuse to announce if less than 90 people are present. Or, if you know you need to reach at least 100 people with two announcements, you could choose never to announce to less than 50 at once. Obviously those approaches risk never announcing at all if the actual data does not meet what you would expect. But that's always going to be a risk, since you could get 1 person in the first group and then 0 in all of the rest, no matter what you do.
Or, you could try more clearly defining the problem, I have a hard time figuring out how to relate this to computers.
Lets start my trying to solve the simplest possible variant of the problem: Lets assume N people and K timeslots, but only one possible announcement. Lets also assume that each person will only ever stay for one timeslot and that each person who hasn't yet shown up has an equally probable chance of showing up at any future timeslot.
Given these simplifications, at each timeslot you look at the payoff of announcing at the current timeslot and compare to the chance of a future timeslot having a higher payoff, eg, lets assume 4 people 3 timeslots:
Timeslot 1: Person 1 shows up, so you know you could get a payoff of 1 by announcing, but then you have 3 people to show up in 2 remaining timeslots, so at least one of those timeslots is guaranteed to have 2 people, so don't announce..
So at each timeslot, you can calculate the chance that a later timeslot will have a higher payoff than the current by treating the remaining (N) people and (K) timeslots as being N independent random numbers each from 1..k, and calculate the chance of at least one value k being hit more than or equal to the current-payoff times. (Similar to the Birthday problem, but for more than 1 collision) and then you need to decide hwo much to discount based on expected variances. (bird in the hand, etc)
Generalization of this solution to the original problem is left as an exercise for the reader.

Resources