Why Mahout doesn't yet have Linear Regression - hadoop

I am just starting to work with Mahout, and one thing which perplexed me a great deal is the lack of Linear Regression. Even logistic regression, which is much harder, is supported to some degree with research going on, but it's all silent on linear regression front!
From what I understand, OLS is one of the easiest problems to solve -
Y = Xb + e
has a linear regression solution of b = (X^T X)^(-1) X^T Y, where X^T is transpose of X, and if the matrix (X^T X) turns out singular (i.e. not invertible) then it's perfectly fine to show error message even though a solution using generalized inverse exists.
Computation of both X^T X and X^ Y are just computations of sums and sum of products of elements, which is probably the easiest thing to do with MapReduce as I understand.
(Which makes me think... is there any module that supports native matrix operations required to compute regression cofficients? That would make a regression module unnecessary indeed...)
Am I missing something which makes regression hard to compute in Mahout?

I don't know if there's a "why" to things like this. It just doesn't exist.
However I think it's the opposite of what you suppose; it's too "easy". Unless you're solving a solution of a ten million equations, it's probably not of a scale that Hadoop is called for. There are plenty of existing packages that can do this really well on one machine. If you want something also in Java from Apache just look at Commons Math for example.
Not to say there couldn't be a fine non-distributed version in the project, but since the emphasis is mostly big-scale and Hadoop, that's probably "why".

I think it's simply because NxN matrix inversion's complexity is O(N^3) and subject to numerical instability, which is quite common whith sparse high-dimensionnal matrices.
Does anyone have another explanation or can someone confirm my thoughts?

Related

Python: How to solve DAE with Jacobian efficiently?

I am trying to use the Assimulo package to solve a set of differential algebraic equations (DAEs). I am trying to use an algorithm similar to that shown here. However, there does not seem to be an option to pass in a sparse matrix. My Jacobian matrix is very large, approximately 3000 x 3000. Do you know if there is a way to solve my DAEs more computationally efficiently?
In my experience with sparse ODE systems (more precisely with systems of semi-discretized PDEs), using an iterative linear solver greatly enhances numerical efficiency. As far as I know, Assimulo doesn't allow to provide a jacobian sparsity pattern, but changing the linear solver is another way to tackle this.
You would do something like:
model = Explicit_Problem(ode_function, y0=y_init, t0=t_init)
simulator = CVode(model)
sim.linear_solver = 'SPGMR'
I'm not sure if this also applies for DAE systems, but I think it's worth giving this a try.

Implementing a Least Squares Kernel classifier

I am trying to find the equation I would need to use in order to implement a Least Squares Kernel classifier for a dataset with N samples of feature length d. I have the kernel equation k(x_i, x_j) and I need the equation to pug it into to get the length-d vector used to classify future data. No matter where I look/google, Although there are dozens of powerpoints and pdfs that seem to give me almost what I'm looking for, I can't find a resource which can give me a straight answer.
note: I am not looking for the programming-language tool that computes this for me such as lsqlin, but the mathematical formula.
Least Squares Kernel SVM (what I assume your actually asking about) is equivalent to Kernelized Ridge Regression. This is the simplest what to implement it, and the solution can be found here, assume you have the appropriate background.

Which optimization algorithm should I use to optimize the weights of a multilayer perceptron?

Actually these are 3 questions:
Which optimization algorithm should I use to optimize the weights of a multilayer perceptron, if I knew...
1) only the value of the error function? (blackbox)
2) the gradient? (first derivative)
3) the gradient and the hessian? (second derivative)
I heard CMA-ES should work very well for 1) and BFGS for 2) but I would like to know if there are any alternatives and I don't know wich algorithm to take for 3).
Ok, so this doesn't really answer the question you initially asked, but it does provide a solution to the problem you mentioned in the comments.
Problems like dealing with a continuous action space are normally not dealt with via changing the error measure, but rather by changing the architecture of the overall network. This allows you to keep using the same highly informative error information while still solving the problem you want to solve.
Some possible architectural changes that could accomplish this are discussed in the solutions to this question. In my opinion, I'd suggest using a modified Q-learning technique where the state and action spaces are both represented by self organizing maps, which is discussed in a paper mentioned in the above link.
I hope this helps.
I solved this problem finally: there are some efficient algorithms for optimizing neural networks in reinforcement learning (with fixed topology), e. g. CMA-ES (CMA-NeuroES) or CoSyNE.
The best optimization algorithm for supervised learning seems to be Levenberg-Marquardt (LMA). This is an algorithm that is specifically designed for least square problems. When there are many connections and weights, LMA does not work very well because the required space is huge. In this case I am using Conjugate Gradient (CG).
The hessian matrix does not accelerate optimization. Algorithms that approximate the 2nd derivative are faster and more efficient (BFGS, CG, LMA).
edit: For large scale learning problems often Stochastic Gradient Descent (SGD) outperforms all other algorithms.

Performance Testing for Calculation-Heavy Programs

What are some good tips and/or techniques for optimizing and improving the performance of calculation heavy programs. I'm talking about things like complication graphics calculations or mathematical and simulation types of programming where every second saved is useful, as opposed to IO heavy programs where only a certain amount of speedup is helpful.
While changing the algorithm is frequently mentioned as the most effective method here,I'm trying to find out how effective different algorithms are in the first place, so I want to create as much efficiency with each algorithm as is possible. The "problem" I'm solving isn't something thats well known, so there are few if any algorithms on the web, but I'm looking for any good advice on how to proceed and what to look for.
I am exploring the differences in effectiveness between evolutionary algorithms and more straightforward approaches for a particular group of related problems. I have written three evolutionary algorithms for the problem already and now I have written an brute force technique that I am trying to make as fast as possible.
Edit: To specify a bit more. I am using C# and my algorithms all revolve around calculating and solving constraint type problems for expressions (using expression trees). By expressions I mean things like x^2 + 4 or anything else like that which would be parsed into an expression tree. My algorithms all create and manipulate these trees to try to find better approximations. But I wanted to put the question out there in a general way in case it would help anyone else.
I am trying to find out if it is possible to write a useful evolutionary algorithm for finding expressions that are a good approximation for various properties. Both because I want to know what a good approximation would be and to see how the evolutionary stuff compares to traditional methods.
It's pretty much the same process as any other optimization: profile, experiment, benchmark, repeat.
First you have to figure out what sections of your code are taking up the time. Then try different methods to speed them up (trying methods based on merit would be a better idea than trying things at random). Benchmark to find out if you actually did speed them up. If you did, replace the old method with the new one. Profile again.
I would recommend against a brute force approach if it's at all possible to do it some other way. But, here are some guidelines that should help you speed your code up either way.
There are many, many different optimizations you could apply to your code, but before you do anything, you should profile to figure out where the bottleneck is. Here are some profilers that should give you a good idea about where the hot spots are in your code:
GProf
PerfMon2
OProfile
HPCToolkit
These all use sampling to get their data, so the overhead of running them with your code should be minimal. Only GProf requires that you recompile your code. Also, the last three let you do both time and hardware performance counter profiles, so once you do a time (or CPU cycle) profile, you can zoom in on the hotter regions and find out why they might be running slow (cache misses, FP instruction counts, etc.).
Beyond that, it's a matter of thinking about how best to restructure your code, and this depends on what the problem is. It may be that you've just got a loop that the compiler doesn't optimize well, and you can inline or move things in/out of the loop to help the compiler out. Or, if you're running as fast as you can with basic arithmetic ops, you may want to try to exploit vector instructions (SSE, etc.) If your code is parallel, you might have load balance problems, and you may need to restructure your code so that data is better distributed across cores.
These are just a few examples. Performance optimization is complex, and it might not help you nearly enough if you're doing a brute force approach to begin with.
For more information on ways people have optimized things, there were some pretty good examples in the recent Why do you program in assembly? question.
If your optimization problem is (quasi-)convex or can be transformed into such a form, there are far more efficient algorithms than evolutionary search.
If you have large matrices, pay attention to your linear algebra routines. The right algorithm can make shave an order of magnitude off the computation time, especially if your matrices are sparse.
Think about how data is loaded into memory. Even when you think you're spending most of your time on pure arithmetic, you're actually spending a lot of time moving things between levels of cache etc. Do as much as you can with the data while it's in the fastest memory.
Try to avoid unnecessary memory allocation and de-allocation. Here's where it can make sense to back away from a purely OO approach.
This is more of a tip to find holes in the algorithm itself...
To realize maximum performance, simplify everything inside the most inner loop at the expense of everything else.
One example of keeping things simple is the classic bouncing ball animation. You can implement gravity by looking up the definition in your physics book and plugging in the numbers, or you can do it like this and save precious clock cycles:
initialize:
float y = 0; // y coordinate
float yi = 0; // incremental variable
loop:
y += yi;
yi += 0.001;
if (y > 10)
yi = -yi;
But now let's say you're having to do this with nested loops in an N-body simulation where every particle is attracted to every other particle. This can be an enormously processor intensive task when you're dealing with thousands of particles.
You should of course take the same approach as to simplify everything inside the most inner loop. But more than that, at the very simplest level you should also use data types wisely. For example, math operations are faster when working with integers than floating point variables. Also, addition is faster than multiplication, and multiplication is faster than division.
So with all of that in mind, you should be able to simplify the most inner loop using primarily addition and multiplication of integers. And then any scaling down you might need to do can be done afterwards. To take the y and yi example, if yi is an integer that you modify inside the inner loop then you could scale it down after the loop like this:
y += yi * 0.01;
These are very basic low-level performance tips, but they're all things I try to keep in mind whenever I'm working with processor intensive algorithms. Of course, if you then take these ideas and apply them to parallel processing on a GPU then you can take your algorithm to a whole new level. =)
Well how you do this depends the most on which language
you are using. Still, the key in any language
in the profiler. Profile your code. See which
functions/operations are taking the most time and then determine
if you can make these costly operations more efficient.
Standard bottlenecks in numerical algorithms are memory
usage (do you access matrices in the order which the elements
are stored in memory); communication overhead, etc. They
can be little different than other non-numerical programs.
Moreover, many other factors such as preconditioning, etc.
can lead to drastically difference performance behavior
of the SAME algorithm on the same problem. Make sure
you determine optimal parameters for your implementations.
As for comparing different algorithms, I recommend
reading the paper
"Benchmarking optimization software with performance profiles,"
Jorge Moré and Elizabeth D. Dolan, Mathematical Programming 91 (2002), 201-213.
It provides a nice, uniform way to compare different algorithms being
applied to the same problem set. It really should be better known
outside of the optimization community (in my not so humble opinion
at least).
Good luck!

Efficiency/speed for trigonometric functions

In a game I'm making, I've got two points, pt1 and pt2, and I want to work out the angle between them. I've already worked out the distance, in an earlier calculation. The obvious way would be to arctan the horizontal distance over the vertical distance (tan(theta) = opp/adj).
I'm wondering though, as I've already calculated the distance, would it be quicker to use arcsine/arccosine with the distance and dx or dy?
Also, might I be better off pre-calculating in a table?
I suspect there's a risk of premature optimization here. Also, be careful about your geometry. Your opposite/adjacent approach is a property of right angle triangles, is that what you actually have?
I'm assuming your points are planar, and so for the general case you have them implicitly representing two vectors form the origin (call these v1 v2), so your angle is
theta=arccos(dot(v1,v2)/(|v1||v2|)) where |.| is vector length.
Making this faster (assuming the need) will depend on a lot of things. Do you know the vector lengths, or have to compute them? How fast can you do a dot product in your architecture. How fast is acos? At some point tricks like table lookup (probably interpolated) might help but that will cost you accuracy.
It's all trade-offs though, there really isn't a general answer to your question.
[edit: added commentary]
I'd like to re-emphasize that often playing "x is fastest" is a bit of a mugs game with modern cpus and compilers anyway. You won't know until you measure it and grovel the generated code. When you hit the point that you really care about it at this level for a (hopefully small) piece of code, you can find out in detail what your system is doing. But it's painstaking. Maybe a table is good. But maybe you've got fast vector computations and a small cache. etc. etc. etc. It all amounts to "it depends". Sorry 'bout that. On the other hand, if you haven't reached the point that you really care so much about this bit of code... you probably shouldn't be thinking about it at this level at all. Make it right. Make it clean (which means abstraction as well as code). Then worry about the overhead.
Aside from all of the wise comments regarding premature optimization, let's just assume this is the hotspot and do a frigg'n benchmark:
Times are in nanoseconds, scaled to normalize 'acos' between the systems.
'acos' simply assumes unit radius i.e. acos(adj), whereas 'acos+div' means acos(adj/hyp).
System 1 is a 2.4GHz i5 running Mac OS X 10.6.4 (gcc 4.2.1)
System 2 is a 2.83GHz Core2 Quad running Red Hat 7 Linux 2.6.28 (gcc 4.1.2)
System 3 is a 1.66GHz Atom N280 running Ubuntu 10.04 2.6.32 (gcc 4.4.3)
System 4 is a 2.40GHz Pentium 4 running Ubuntu 10.04 2.6.32 (gcc 4.4.3)
Summary: Relative performance is all over the map. Sometimes atan2 is faster, sometimes its slower. Very strangely, on some systems doing acos with a division is faster than doing it without. Test on your own system :-/
If you're going to be doing this many times, pre-calculate in a table. Performance will be much better this way.
Tons of good answers here.
By the way, if you use Math.atan2, you get a full 2π of angles out of it.
I would just do it, then run it flat out. If you don't like the speed, and if samples show that you're actually in that code most of the time and not someplace else,
try replacing it with table lookup. If you don't need precision closer than 1 degree, you could use a pretty small table and interpolation.
Also, you may want to memoize the function. Why recompute something you already did recently?
Added: If you use a table, it only has to cover angles from 0-45 degrees (and it can be hard-coded). You can get everything else by symmetry.
From a pure speed standpoint, a precalculated table and a closest-match lookup would be best. It involves some overhead, of course, depending on how fine-grained you need the angle to be, but it's more than worth it if you're doing this calculation a lot (or in a tight loop), as those are going to be expensive calculations.
Get it right first !
And then profile and optimize. Table lookup is a good candidate for sure, but be sure to have your calculation right before doing anything fancy
If you're interested in big-O notation, all the methods you might use are O(1).
If you're interested in what works fastest, test it. Write a wrapper function, one that calls your preferred method but can be easily changed, and test with that. Make sure that your application spends a noticeable amount of time doing this, so you aren't wasting your own time. Try whatever ways occur to you. Ideally, run it on more than one different CPU.
I've become very leery of predicting what will take more or less time on modern processors. Lookup tables used to be the answer if you needed speed, but you don't know a priori the effects on caching or how long it's going to take to normalize and look up versus how long it's going to take to do a trig function on a particular CPU.
Given that this is for a game, you probably care about speed. A lookup table is definitely the fastest but you trade accuracy for speed with this method. So how accurate must you be to meet requirements? Only you can answer that. Before you trade accuracy, determine first if you have a speed problem. All of the trigonometric functions are calculated using numerical methods (research numerical analysis to learn more). Some trig functions are have more expensive methods than others because they rely on series that converge more slowly and who knows, your computer may have different implementations for these functions than another computer. At any rate, you can find out for yourself how expensive these functions are by writing some small programs that loop through as many iterations as you desire, with increments of your choosing, all the while timing the outcomes. Then you can pick the fastest method.
While others are very right to mention that you are almost certainly falling into the pit of premature optimization, when they say that trigonometric functions are O(1) they're not telling the whole story.
Most trigonometric function implementations are actually O(N) in the value of the input function. This is because the trig functions are most efficiently calculated on a small interval like [0, 2π) (or, for the best implementations, even smaller parts of this interval, but that one suffices to explain things). So the algorithm looks something like this, in pseudo-Python:
def Cosine_0to2Pi(x):
#a series approximation of some kind, or CORDIC, or perhaps a table
#this function requires 0 <= x < 2Pi
def MyCosine(x):
if x < 0:
x = -x
while x >= TwoPi:
x -= TwoPi
return Cosine_0to2Pi(x)
Even microcoded CPU instructions like the x87's FSINCOS end up doing something like this internally. So trig functions, because they are periodic, usually take O(N) time to do the argument reduction. There are two caveats, however:
If you have to calculate a ton of values off the principal domain of the trig functions, your math is probably not very well thought out.
Big-O notation hides a constant factor. Argument reduction has a very small constant factor, because it's simple to do. Thus the O(1) part is going to dominate the O(N) part for just about every input.

Resources