Here's a theoretical question. Let's assume that I have implemented two types of collaborative filtering: user-based CF and item-based CF (in the form of Slope One).
I have a nice data set for these algorithms to run on. But then I want to do two things:
I'd like to add a new rating to the data set.
I'd like to edit an existing rating.
How should my algorithms handle these changes (without doing a lot of unnecessary work)? Can anyone help me with that?
For both cases, the strategy is very similar:
user-based CF:
update all similarities for the affected user (that is, one row and one column in the similarity matrix)
if your neighbors are precomputed, compute the neighbors for the affected user (for a complete update, you may have to recompute all neighbors, but I would stick with the approximate solution)
Slope-One:
update the frequency (only in the 'add' case) and the diff matrix entries for the affected item (again, one row and one column)
Remark: If your 'similarity' is asymmetric, you need to update one row and one column. If it is symmetric, updating one row automatically results in updating the corresponding column.
For Slope-One, the matrices are symmetric (frequency) and skew symmetric (diffs), so if you handle you also need to update one row or column, and get the other one for free (if your matrix storage works like this).
If you want to see an example of how this could be implemented, have a look at MyMediaLite (disclaimer: I am the main author): https://github.com/zenogantner/MyMediaLite/blob/master/src/MyMediaLite/RatingPrediction/ItemKNN.cs
The interesting code is in the method RetrainItem(), which is called from AddRatings() and UpdateRatings().
The general thing are called online algorithms.
Instead of retraining the whole predictor, it can be updated "online" (while remaining useable) with the new data only.
If you google for "online slope one predictor" you should be able to find some relevant approaches from literature.
Related
I'm trying to implement this paper
Browser Fingerprint Coding Methods Increasing the Effectiveness of User Identification in the Web Traffic
I got a couple of questions about the LHS algorithm in general and the proposed implementation:
The LSH algorithm it's used only when you have a lot of documents to compare with each other (because it is supposed to put the similar ones in the same bucket from what I got). If for example I have a new document and I want to calculate the similarity with the others, I have to relaunch the LHS algorithm from scratch, including the new document, correct?
In 'Mining of Massive Datasets, Ch3', it is said that for the LHS we should use one hash function per band. Each hash function creates n buckets.
So, for the first band, we are going to have n buckets. For the second band onward, Am I supposed to keep using the same hash function (so this way I keep using the same buckets as before) or another one (ending so with m>>n buckets)?
This question is related t the previous one. If I use the same hash function for all the bands, then I'll have n buckets. No problem here. But If I have to use more hash functions (one different function per row), I'm going to end up with a lot of different buckets. Am I supposed to measure the similarity for each pair in each bucket? (If I have to use only one hash function then here it's not a problem).
In the paper, I understood most of the algorithm except for its end.
Basically, two Signatures matrices are created (one for stable features and one for unstable features) via minhashing. Then, they use LSH on the first matrix to obtain a list of candidates pairs. So far so good.
What happens at the end? do they perform the LHS on the second matrix? How the result of the first LHS is used? I cannot see the relationship between the first and the second LHS.
The output of the final step is supposed to be a list of pairing candidates, right? and all that I have to do is performing Jaccard similarity on them and setting a threshold, right?
Thanks for your answers!
I got a partial answer to my question (still missing question 4)
No. You would keep the bucket structure and hash the new doc into it. Then compare with only those docs in one of the buckets it fell into.
No. You HAVE to use different hash functions and a different set of buckets for each hash function.
This is irrelevant because of the answer to (2).
\ am dealing with a matrix in MATLAB which is sparse and has many rows and columns. In this case, the row and columns of the matrix are the ids for particular items. Let's assume them as id1 and id2.
It would be nice if the ids for rows and columns could be embedded so I can have access to them easily to them without the need for creating extra variables that keep the two ids.
The answer would be probably to use a table data type. Tables are very ideal answer for my need however I was wondering if I could create a table data type for a sparse matrix?
A [m*n] sparse matrix %% m & n are huge
id1 [1*m] , id2 [1*n] %% two vectors containing numeric ids for rows and column
Could we obtain?
T [m*n] sparse table matrix
Thanks for sharing your view with me.
I will address the question and the comments in order to clear some confusion.
The short answer
There is no sparse table class in Matlab. Cannot do. Use sparse() matrices.
The long answer
There is a reason why sparse tables make little sense:
Philosophically speaking, the advantage of having nice row and column labels, is completely lost if you are working with a big panel of data and/or if the data is sparse.
Scrolling through 246829 rows and 33336 columns? Can only be useful at very isolated times if you are debugging your code and a specific outlier is causing you results to go off. Also, all you might see is just a sea of zeros.
Technically a table can have more columns for the same variable, i.e. table(rand(10,2), rand(10,1)) is a valid table. How would you consider define sparsity on such table?
Fine, suppose you are working with a matrix-like table, i.e. one element per table cell and same numeric class. Still, none of the algebraic operators are defined on a table(). So you need to extract the content first, in order to be able to perform any operation that spans more than a single column of data. Just to be clear, once the data is extracted, then you have e.g. your double (full) matrix or in an ideal case a double sparse matrix.
Now, a few misconceptions to clear:
Less variables implies clearer/cleaner code. Not true. You are probably thinking about the extreme case (in bad practices) of how do I make a series of variables a1, a2, a3, etc..
There is a sweet spot between verbosity and number of variables, amount of comments, and code clarity/maintainability. Only with time and experience you find the right balance.
Control over data cannot go without visual inspection. This approach does NOT scale with big data and the sooner you abandon it, the faster your code will become more reliable. You need to verify your results systematically, rather than relying on visual inspection. Failure to (visually) spot a problem in the data, grows exponentially with its dimension, faster than with systematic tests.
Some background info on my work:
I work with high-frequency prices, that's terabytes of data. I also extended the table() class with additional methods and fixes to help me with my work (see https://github.com/okomarov/tableutils), but I do not see how sparsity is a useful feature to add to table().
I have two (or three) classes and each classes can only possess one label.
I want to optimize (automatically if possible) parameters and thresholds of classifiers in order for my first class to contain only 100 % sure data. Even if it contains a small number of instances.
I don't mind for the remaining classes to contain false alarm or correct rejection.
I don't mind to have unclassified data.
I have already been searching on stackoverflow and on the weka's wiki but maybe my lack of knowledge concerning weka made me miss some keywords.
I also tried to perform the task with the well-known "iris" database but I think that in this case, any class can be 100 % sure.
Yet, I have only succeed in testing multiple classifiers and tuning them manually but without performing 100 % correct for my first class. (I checked this result in the confusion matrix given by weka's report.)
Somehow, I know it is possible for my class to contain 100% sure data because I managed to do it in Matlab with simple threshold set manually. But I would like to try out a bigger database, to obtain better threshold and to use the power of weka.
Any suggestions would be helpful, thanks !
You probably need the "Cost Sensitive Classifier" among "meta" classifiers.
If you are working in the Explorer, here is the dialog you get.
Choose the your "classifier" (something beyond ZeroR :) ).
Set your "cost matrix". For 2-class problem this will be 2x2 matrix.
By setting one non-diagonal component very large (>>1, let us say 1000) you ensure that misclassifying one class (your "first" class) is 1000 times more expensive than misclassifying another class. This should do the job.
New with Matlab.
When I try to load my own date using the NN pattern recognition app window, I can load the source data, but not the target (it is never on the drop down list). Both source and target are in the same directory. Source is 5000 observations with 400 vars per observation and target can take on 10 different values (recognizing digits). Any Ideas?
Before you do anything with your own data you might want to try out the example data sets available in the toolbox. That should make many problems easier to find later on because they definitely work, so you can see what's wrong with your code.
Regarding your actual question: Without more details, e.g. what your matrices contain and what their dimensions are, it's hard to help you. In your case some of the problems mentioned here might be similar to yours:
http://www.mathworks.com/matlabcentral/answers/17531-problem-with-targets-in-nprtool
From what I understand about nprtool your targets have to consist of a matrix with only one 1 (for the correct class) in either row or column (depending on the input matrix), so make sure that's the case.
Why is Shannon's Entropy measure used in Decision Tree branching?
Entropy(S) = - p(+)log( p(+) ) - p(-)log( p(-) )
I know it is a measure of the no. of bits needed to encode information; the more uniform the distribution, the more the entropy. But I don't see why it is so frequently applied in creating decision trees (choosing a branch point).
Because you want to ask the question that will give you the most information. The goal is to minimize the number of decisions/questions/branches in the tree, so you start with the question that will give you the most information and then use the following questions to fill in the details.
For the sake of decision trees, forget about the number of bits and just focus on the formula itself. Consider a binary (+/-) classification task where you have an equal number of + and - examples in your training data. Initially, the entropy will be 1 since p(+) = p(-) = 0.5. You want to split the data on an attribute that most decreases the entropy (i.e., makes the distribution of classes least random). If you choose an attribute, A1, that is completely unrelated to the classes, then the entropy will still be 1 after splitting the data by the values of A1, so there is no reduction in entropy. Now suppose another attribute, A2, perfectly separates the classes (e.g., the class is always + for A2="yes" and always - for A2="no". In this case, the entropy is zero, which is the ideal case.
In practical cases, attributes don't typically perfectly categorize the data (the entropy is greater than zero). So you choose the attribute that "best" categorizes the data (provides the greatest reduction in entropy). Once the data are separated in this manner, another attribute is selected for each of the branches from the first split in a similar manner to further reduce the entropy along that branch. This process is continued to construct the tree.
You seem to have an understanding of the math behind the method, but here is a simple example that might give you some intuition behind why this method is used: Imagine you are in a classroom that is occupied by 100 students. Each student is sitting at a desk, and the desks are organized such there are 10 rows and 10 columns. 1 out of the 100 students has a prize that you can have, but you must guess which student it is to get the prize. The catch is that everytime you guess, the prize is decremented in value. You could start by asking each student individually whether or not they have the prize. However, initially, you only have a 1/100 chance of guessing correctly, and it is likely that by the time you find the prize it will be worthless (think of every guess as a branch in your decision tree). Instead, you could ask broad questions that dramatically reduce the search space with each question. For example "Is the student somewhere in rows 1 though 5?" Whether the answer is "Yes" or "No" you have reduced the number of potential branches in your tree by half.