We Keep Coding

NEAT algorithm result precision

I am a PhD student who is trying to use the NEAT algorithm as a controller for a robot and I am having some accuracy issues with it. I am working with Python 2.7 and for it and am using two NEAT python implementations:
The NEAT which is in this GitHub repository: https://github.com/CodeReclaimers/neat-python
Searching in Google, it looks like it has been used in some projects with succed.
The multiNEAT library developed by Peter Chervenski and Shane Ryan: http://www.multineat.com/index.html.
Which appears in the "official" software web page of NEAT software catalog.
While testing the first one, I've found that my program converges quickly to a solution, but this solution is not precise enough. As lack of precision I want to say a deviation of a minimum of 3-5% in the median and average related to the "perfect" solution at the end of the evolution (Depending on the complexity of the problem, an error around 10% is normal for my solutions. Furthermore, I could said that I've "never" seen an error value under the 1% between the solution given by the NEAT and the solution that it is the correct one). I must said that I've tried a lot of different parameter combinations and configurations (this is an old problem for me).
Due to that, I tested the second library. The MultiNEAT library converges quickly and easier that the previous one. (I assume that is due to the C++ implementation instead the pure Python) I get similar results, but I still have the same problem; lack of accuracy. This second library has different configuration parameters too, and I haven't found a proper combination of them to improve the performance of the problem.
My question is:
Is it normal to have this lack of accuracy in the NEAT results? It achieves good solutions, but not good enough for controlling a robot arm, which is what I want to use it for.
I'll write what I am doing in case someone sees some conceptual or technical mistake in the way I set out my problem:
To simplify the problem, I'll show a very simple example: I have a very simple problem to solve, I want a NN that may calculate the following function: y = x^2 (similar results are found with y=x^3 or y = x^2 + x^3 or similar functions)
The steps that I follow to develop the program are:
"Y" are the inputs to the network and "X" the outputs. The
activation functions of the neural net are sigmoid functions.
I create a data set of "n" samples given values to "X" between the
xmin = 0.0 and the xmax = 10.0
As I am using sigmoid functions, I make a normalization of the "Y"
and "X" values:
"Y" is normalized linearly between (Ymin, Ymax) and (-2.0, 2.0) (input range of sigmoid).
"X" is normalized linearly between (Xmin, Xmax) and (0.0, 1.0) (the output range of sigmoid).
After creating the data set, I subdivide in in a train sample (70%
percent of the total amount), a validation sample and a test sample
(15% each one).
At this point, I create a population of individuals for doing
evolution. Each individual of the population is evaluated in all the
train samples. Each position is evaluated as:
eval_pos = xmax - abs(xtarget - xobtained)
And the fitness of the individual is the average value of all the train positions (I've selected the minimum too but it gives me worse performance).
After the whole evaluation, I test the best obtained individual
against the test sample. And here is where I obtained those
"un-precise values". Moreover, during the evaluation process, the
maximum value where "abs(xtarget - xobtained) = 0" is never
obtained.
Furthermore, I assume that how I manipulate the data is right because, I use the same data set for training a neural network in Keras and I get much better results than with NEAT (an error less than a 1% is achievable after 1000 epochs in a layer with 5 neurons).
At this point, I would like to know if what is happened is normal because I shouldn't use a data set of data for developing the controller, it must be learned "online" and NEAT looks like a suitable solution for my problem.
Thanks in advance.
EDITED POST:
Firstly, Thanks for comment nick.
I'll answer your questions below::
I am using the NEAT algorithm.
Yes, I've carried out experiments increasing the number of individuals in the population and the generations number. A typical graph that I get is like this:
Although the population size in this example is not such big, I've obtained similar results in experiments incrementing the number of individuals or the number of generations. Populations of 500 in individuals and 500 generations, for example. In this experiments, the he algorithm converge fast to a solution, but once there, the best solution is stucked and it does not improve any more.
As I mentioned in my previous post, I've tried several experiments with many different parameters configurations... and the graphics are more or less similar to the previous showed.
Furthermore, other two experiments that I've tried were: once the evolution reach the point where the maximum value and the median converge, I generate other population based on that genome with new configuration parameters where:
The mutation parameters change with a high probability of mutation (weight and neuron probability) in order to find new solutions with the aim to "jumping" from the current genome to other better.
The neuron mutation is reduced to 0, while the weight "mutation probability" increase for "mutate weight" in a lower range in order to get slightly modifications with the aim to get a better adjustment of the weights. (trying to get a "similar" functionality as backprop. making slighty changes in the weights)
This two experiments didn't work as I expected and the best genome of the population was also the same of the previous population.
I am sorry, but I do not understand very well what do you want to say with "applying your own weighted penalties and rewards in your fitness function". What do you mean with including weight penalities in the fitness function?
Regards!

Disclaimer: I have contributed to these libraries.
Have you tried increasing the population size to speed up the search and increasing the number of generations? I use it for a trading task, and by increasing the population size my champions were found much sooner.
Another thing to think about is applying your own weighted penalties and rewards in your fitness function, so that anything that doesn't get very close right away is "killed off" sooner and the correct genome is found faster. It should be noted that neat uses a fitness function to learn as a opposed to gradient descent so it wont converge in the same way and its possible you may have to train a bit longer.
Last question, are you using the neat or hyperneat algo from multineat?

Appropriate encoding using Particle Swarm Optimization

The Problem
I've been doing a bit of research on Particle Swarm Optimization, so I said I'd put it to the test.
The problem I'm trying to solve is the Balanced Partition Problem - or reduced simply to the Subset Sum Problem (where the sum is half of all the numbers).
It seems the generic formula for updating velocities for particles is
but I won't go into too much detail for this question.
Since there's no PSO attempt online for the Subset Sum Problem, I looked at the Travelling Salesman Problem instead.
They're approach for updating velocities involved taking sets of visited towns, subtracting one from another and doing some manipulation on that.
I saw no relation between that and the formula above.
My Approach
So I scrapped the formula and tried my own approach to the Subset Sum Problem.
I basically used gbest and pbest to determine the probability of removing or adding a particular element to the subset.
i.e - if my problem space is [1,2,3,4,5] (target is 7 or 8), and my current particle (subset) has [1,None,3,None,None], and the gbest is [None,2,3,None,None] then there is a higher probability of keeping 3, adding 2 and removing 1, based on gbest
I can post code but don't think it's necessary, you get the idea (I'm using python btw - hence None).
So basically, this worked to an extent, I got decent solutions out but it was very slow on larger data sets and values.
My Question
Am I encoding the problem and updating the particle "velocities" in a smart way?
Is there a way to determine if this will converge correctly?
Is there a resource I can use to learn how to create convergent "update" formulas for specific problem spaces?
Thanks a lot in advance!

Encoding
Yes, you're encoding this correctly: each of your bit-maps (that's effectively what your 5-element lists are) is a particle.
Concept
Your conceptual problem with the equation is because your problem space is a discrete lattice graph, which doesn't lend itself immediately to the update step. For instance, if you want to get a finer granularity by adjusting your learning rate, you'd generally reduce it by some small factor (say, 3). In this space, what does it mean to take steps only 1/3 as large? That's why you have problems.
The main possibility I see is to create 3x as many particles, but then have the transition probabilities all divided by 3. This still doesn't satisfy very well, but it does simulate the process somewhat decently.
Discrete Steps
If you have a very large graph, where a high velocity could give you dozens of transitions in one step, you can utilize a smoother distance (loss or error) function to guide your model. With something this small, where you have no more than 5 steps between any two positions, it's hard to work with such a concept.
Instead, you utilize an error function based on the estimated distance to the solution. The easy one is to subtract the particle's total from the nearer of 7 or 8. A harder one is to estimate distance based on that difference and the particle elements "in play".
Proof of Convergence
Yes, there is a way to do it, but it requires some functional analysis. In general, you want to demonstrate that the error function is convex over the particle space. In other words, you'd have to prove that your error function is a reliable distance metric, at least as far as relative placement goes (i.e. prove that a lower error does imply you're closer to a solution).
Creating update formulae
No, this is a heuristic field, based on shape of the problem space as defined by the particle coordinates, the error function, and the movement characteristics.
Extra recommendation
Your current allowable transitions are "add" and "delete" element.
Include "swap elements" to this: trade one present member for an absent one. This will allow the trivial error function to define a convex space for you, and you'll converge in very little time.

Find best interpolation nodes

I am studying physics and ran into a really interesting problem. I'm not an expert on programming so please take this into account while reading this.
I really hope that someone can help me with this problem because I struggle with this matter for about 2 months now and don't see any success.
So here is my Problem:
I got a bunch of data sets (more than 2 less than 20) from numerical calculations. The set is given by x against measurement values. I have a set of sensors and want to find the best positions x for my sensors such that the integral of the interpolation comes as close as possible to the integral of the numerical data set.
As this sounds like a typical mathematical problem I started to look for some theorems but I did not find anything.
So I started to write a python program based on the SLSQP minimizer. I chose this because it can handle bounds and constraints. (Note there is always a sensor at 0 and one at 1)
Constraints: the sensor array must stay sorted all the time such that x_i smaller than x_i+1 and the interval of x is normalized to [0,1].
Before doing an overall optimization I started to look for good starting points and searched for maximums, minimums and linear areas of my given data sets. But an optimization over 40 values turned out to deliver bad results.
In my second try I started to search for these points and defined certain areas. So I optimized each area with 1 to 40 sensors. Then I compared the results and decided which area is worth putting more sensors in. I the last step I wanted to do an overall optimization again. But these idea didn't seem to be the proper solution, too, because the optimization had convergence problem as well.
The big problem was, that my optimizer broke the boundaries. I covered this by interrupting the optimization, because once this boundaries were broken the result was not correct in the end. If this happens I reset my initial setup and a homogeneous distribution. After this there are normally no violence of boundaries but the results seems to be a homogeneous distribution, too, often this is obviously not the perfect distribution.
As my algorithm works for simple examples and dies for more complex data I think there is a general problem and not just some error in my coding. Does anyone have an idea how to move on or knows some theoretical stuff about this matter?
The attached plot show the areas in different colors. The function is shown at the bottom and the sensor positions are represented as dots. Dots at value y=1 are from the optimization with one sensors 2 represents the results of optimization with 2 variables. So as the program reaches higher sensor numbers the whole thing gets more and more homogeneous.
It is easy to see that if n is the number of sensors and n goes to infinity you have a total homogeneous distribution. But as far as I see this this should not happen for just 10 sensors.

Scipy optimize: Set maximum error

I'm trying to optimize a 4 dimensional function with scipy. Everything works so far, except that I'm not satisfied with the quality of the solution. Right now I have ground truth data, which I use to verify my code. What I get so far is:
End error: 1.52606896507e-05
End Gradient: [ -1.17291295e-05 2.60362493e-05 5.15347856e-06 -2.72388430e-05]
Ground Truth: [0.07999999..., 0.0178329..., 0.9372903878..., 1.7756283966...]
Reconstructed: [ 0.08375729 0.01226504 1.13730592 0.21389899]
The error itself sounds good, but as the values are totally wrong I want to force the optimization algorithm (BFGS) to do more steps.
In the documentation I found the options 'gtol' and 'norm' and I tried to set both to pretty small values (like 0.0000001) but it did not seem to change anything.
Background:
The problem is, that I try to demodulate waves, so I have sin and cos terms and potentially many local (or global) minima. I use bruteforce search to find a good starting point, witch helps a lot, but it currently seems that the most work is done by that brute force search, as the optimization uses often only one iteration step. So I'm trying to improve that part of the calculation somehow.

Many local minima + hardly any improvement after brute search, that sounds bad. It's hard to say something very specific with the level of detail you provide in the question, so here are vague ideas to try (basically, what I'd do if I suspect my minimizer gets stuck):
try manually starting the minimizer from a bunch of different initial guesses.
try using a stochastic minimizer. You're tagging a question scipy, so try basinhopping
if worst comes to worst, just throw random points in a loop, leave it to work over the lunch break (or overnight)
Also, waves, sines and cosines --- it might be useful to think if you can reformulate your problem in the Fourier space.

I found out that the gradient at the starting point is already very flat (values in 10^-5), so I tried to scale the gradient function witch I already provided. This seemed to be pretty effective, I could force the Algorithm to do much more steps and my results are far better now.
They are not perfect though, but a complete discussion of this is outside of the bounds of this question, so I might start a new one, where I describe the whole problem from bottom up.

Comparing Root-finding (of a function) algorithms in Python

I would like to compare different methods of finding roots of functions in python (like Newton's methods or other simple calc based methods). I don't think I will have too much trouble writing the algorithms
What would be a good way to make the actual comparison? I read up a little bit about Big-O. Would this be the way to go?

The answer from #sarnold is right -- it doesn't make sense to do a Big-Oh analysis.
The principal differences between root finding algorithms are:
rate of convergence (number of iterations)
computational effort per iteration
what is required as input (i.e. do you need to know the first derivative, do you need to set lo/hi limits for bisection, etc.)
what functions it works well on (i.e. works fine on polynomials but fails on functions with poles)
what assumptions does it make about the function (i.e. a continuous first derivative or being analytic, etc)
how simple the method is to implement
I think you will find that each of the methods has some good qualities, some bad qualities, and a set of situations where it is the most appropriate choice.

Big O notation is ideal for expressing the asymptotic behavior of algorithms as the inputs to the algorithms "increase". This is probably not a great measure for root finding algorithms.
Instead, I would think the number of iterations required to bring the actual error below some epsilon ε would be a better measure. Another measure would be the number of iterations required to bring the difference between successive iterations below some epsilon ε. (The difference between successive iterations is probably a better choice if you don't have exact root values at hand for your inputs. You would use a criteria such as successive differences to know when to terminate your root finders in practice, so you could or should use them here, too.)
While you can characterize the number of iterations required for different algorithms by the ratios between them (one algorithm may take roughly ten times more iterations to reach the same precision as another), there often isn't "growth" in the iterations as inputs change.
Of course, if your algorithms take more iterations with "larger" inputs, then Big O notation makes sense.

Big-O notation is designed to describe how an alogorithm behaves in the limit, as n goes to infinity. This is a much easier thing to work with in a theoretical study than in a practical experiment. I would pick things to study that you can easily measure that and that people care about, such as accuracy and computer resources (time/memory) consumed.
When you write and run a computer program to compare two algorithms, you are performing a scientific experiment, just like somebody who measures the speed of light, or somebody who compares the death rates of smokers and non-smokers, and many of the same factors apply.
Try and choose an example problem or problems to solve that is representative, or at least interesting to you, because your results may not generalise to sitations you have not actually tested. You may be able to increase the range of situations to which your results reply if you sample at random from a large set of possible problems and find that all your random samples behave in much the same way, or at least follow much the same trend. You can have unexpected results even when the theoretical studies show that there should be a nice n log n trend, because theoretical studies rarely account for suddenly running out of cache, or out of memory, or usually even for things like integer overflow.
Be alert for sources of error, and try to minimise them, or have them apply to the same extent to all the things you are comparing. Of course you want to use exactly the same input data for all of the algorithms you are testing. Make multiple runs of each algorithm, and check to see how variable things are - perhaps a few runs are slower because the computer was doing something else at a time. Be aware that caching may make later runs of an algorithm faster, especially if you run them immediately after each other. Which time you want depends on what you decide you are measuring. If you have a lot of I/O to do remember that modern operating systems and computer cache huge amounts of disk I/O in memory. I once ended up powering the computer off and on again after every run, as the only way I could find to be sure that the device I/O cache was flushed.

You can get wildly different answers for the same problem just by changing starting points. Pick an initial guess that's close to the root and Newton's method will give you a result that converges quadratically. Choose another in a different part of the problem space and the root finder will diverge wildly.
What does this say about the algorithm? Good or bad?

I would suggest you to have a look at the following Python root finding demo.
It is a simple code, with some different methods and comparisons between them (in terms of the rate of convergence).
http://www.math-cs.gordon.edu/courses/mat342/python/findroot.py

I just finish a project where comparing bisection, Newton, and secant root finding methods. Since this is a practical case, I don't think you need to use Big-O notation. Big-O notation is more suitable for asymptotic view. What you can do is compare them in term of:
Speed - for example here newton is the fastest if good condition are gathered
Number of iterations - for example here bisection take the most iteration
Accuracy - How often it converge to the right root if there is more than one root, or maybe it doesn't even converge at all.
Input - What information does it need to get started. for example newton need an X0 near the root in order to converge, it also need the first derivative which is not always easy to find.
Other - rounding errors
For the sake of visualization you can store the value of each iteration in arrays and plot them. Use a function you already know the roots.

Although this is a very old post, my 2 cents :)
Once you've decided which algorithmic method to use to compare them (your "evaluation protocol", so to say), then you might be interested in ways to run your challengers on actual datasets.
This tutorial explains how to do it, based on an example (comparing polynomial fitting algorithms on several datasets).
(I'm the author, feel free to provide feedback on the github page!)