I would like to ask if someone know some examples of the Heterogeneous Value Difference Metric (HVDM) distance ? also, i would like to ask if there is an implementation of such metric in R?
I will be grateful if someone can give some useful ressource in such way i could compute this distance manually
This is a very involved subject, which is no doubt why you can't find examples. What worries me about your question is that it is very general, and often a given implementation or use case of this sort of machine learning / data mining may need considerable algorithm tuning to make it effective, because the nature of the data will to some extent dictate how your HVDM is best calculated.
Single dimensional euclidean distance can obviously be calculated by D = a - b. 2D distance is Pythagoras, so D = SQRT((a1-b1)^2+(a2-b2)^2), and when you have N dimensional data D = SQRT((a1-b1)^2+(a2-b2)^2+....+(aN-bN)^2).
So, if you are comparing 2 data sets, a and b, with N numerical values, you can now calculate a distance between them...
Note that the square root is probably usually optional for practical purposes since it affects magnitude, but this is a tuning/performance/optimisation issue... and I'm not sure, but maybe some use cases might be better with it and some without.
Since you say your dataset has nominal values in, this makes it more interesting, as euclidean distance is meaningless for nominal values... How you reconcile that depends on the data, if you can assign numerical data to the nominals, that's good, because you can then calculate a euclidean distance again (e.g. banana = {2,4,6}, apple={4,2,2}, pear={3,3,5}, these numbers being characteristics such as shape, colour, squishiness, for example).
Next problem is that because you have nominal and numerical data which is fundamentally different, you almost certainly need to normalise the nominal and numerical so that one doesn't have an unreasonable weight because of the nature of that data. Also it's possible you might split each numerical data set and calculate 2 distances for each data set comparison... again it's a data dependant decision, or a decision you will make when tuning to get good or even sane performance. Sum the normalised results, or calculate a euclidean distance of them.
Normalising, at its simplest, means dividing by the over all range of the data, so 2 bits of data, both normalised will both be reduced to a value between 0 and 1, thus eliminating irrelevant facts like the magnitude of one bit of data is 10,000 times that of the other. Alternative normalising techniques might be appropriate for your data if it can or does have outliers.
In R, You can find UBL Package that use HVDM as option of Distance, at ENNClassif function.
library(datasets)
data(iris)
summary(iris)
#install.packages("UBL")
library(UBL)
# generate an small imbalanced data set
ir<- iris[-c(95:130), ]
# use HDVM as Distance for numeric and nominal features.
irHVDM <- ENNClassif(Species~., ir, k = 3, dist = "HVDM")
Related
i am working on kmeans clustering .
i have 3d dataset as no.days,frequency,food
->day is normalized by means & std deviation(SD) or better to say Standardization. which gives me range of [-2 to 14]
->for frequency and food which are NOMINAL data in my data sets are normalized by DIVIDE BY MAX ( x/max(x) ) which gives me range [0 to 1]
the problem is that the kmeans only considers the day-axis for grouping since there is obvious gap b/w points in this axis and almost ignores the other two of frequency and food (i think because of negligible gaps in frequency and food dims ).
if i apply the kmeans only on day-axis alone (1D) i get the exact similar result as i applied on 3D(days,frequency,food).
"before, i did x/max(x) as well for days but not acceptable"
so i want to know is there any way to normalize the other two nominal data of frequency and food and we get fair scaling based on DAY-axis.
food => 1,2,3
frequency => 1-36
The point of normalization is not just to get the values small.
The purpose is to have comparable value ranges - something which is really hard for attributes of different units, and may well be impossible for nominal data.
For your kind of data, k-means is probably the worst choice, because k-means relies on continuous values to work. If you have nominal values, it usually gets stuck easily. So my main recommendation is to not use k-means.
For k-means to wprk on your data, a difference of 1 must be the same in every attribute. So 1 day difference = difference between food q and food 2. And because k-means is based on squared errors the difference of food 1 to food 3 is 4x as much as food to food 2.
Unless you have above property, don't use k-means.
You can try to use the Value Difference Metric, VDM (or any variant) to convert pretty much every nominal attribute you encounter to a valid numeric representation. An after that you can just apply standardisation to the whole dataset as usual.
The original definition is here:
http://axon.cs.byu.edu/~randy/jair/wilson1.html
Although it should be easy to find implementations for every common language elsewhere.
N.B. for ordered nominal attributes such as your 'frequency' most of the time it is enough to just represent them as integers.
When we are training our model we usually use MLE to estimate our model. I know it means that the most probable data for such a learned model is our training set. But I'm wondering if its probability match 1 exactly or not?
You almost have it right. The Likelihood of a model (theta) for the observed data (X) is the probability of observing X, given theta:
L(theta|X) = P(X|theta)
For Maximum Likelihood Estimation (MLE), you choose the value of theta that provides the greatest value of P(X|theta). This does not necessarily mean that the observed value of X is the most probable for the MLE estimate of theta. It just means that there is no other value of theta that would provide a higher probability for the observed value of X.
In other words, if T1 is the MLE estimate of theta, and if T2 is any other possible value of theta, then P(X|T1) > P(X|T2). However, there still could be another possible value of the data (Y) different than the observed data (X) such that P(Y|T1) > P(X|T1).
The probability of X for the MLE estimate of theta is not necessarily 1 (and probably never is except for trivial cases). This is expected since X can take multiple values that have non-zero probabilities.
To build on what bogatron said with an example, the parameters learned from MLE are the ones that explain the data you see (and nothing else) the best. And no, the probability is not 1 (except in trivial cases).
As an example (that has been used billions of times) of what MLE does is:
If you have a simple coin-toss problem, and you observe 5 results of coin tosses (H, H, H, T, H) and you do MLE, you will end up giving p(coin_toss == H) a high probability (0.80) because you see Heads way too many times. There are good and bad things about MLE obviously...
Pros: It is an optimization problem, so it is generally quite fast to solve (even if there isn't an analytical solution).
Cons: It can overfit when there isn't a lot of data (like our coin-toss example).
The example I got in my stat classes was as follows:
A suspect is on the run ! Nothing is known about them, except that they're approximatively 1m80 tall. Should the police look for a man or a woman ?
The idea here is that you have a parameter for your model (M/F), and probabilities given that parameter. There are tall men, tall women, short men and short women. However, in the absence of any other information, the probability of a man being 1m80 is larger than the probability of a woman being 1m80. Likelihood (as bogatron very well explained) is a formalisation of that, and maximum likelihood is the estimation method based on favouring parameters which are more likely to result in the actual observations.
But that's just a toy example, with a single binary variable... Let's expand it a bit: I threw two identical die, and the sum of their value is 7. How many side did my die have ? Well, we all know that the probability of two D6 summing to 7 is quite high. But it might as well be D4, D20, D100, ... However, P(7 | 2D6) > P(7 | 2D20), and P(7 | 2D6) > P(7 | 2D100) ..., so you might estimate that my die are 6-faced. That doesn't mean it's true, but its a reasonable estimation, in the absence of any additional information.
That's better, but we're not in machine-learning territory yet... Let's get there: if you want to fit your umptillion-layer neural network on some empirical data, you can consider all possible parameterisations, and how likely each of them is to return the empirical data. That's exploring an umptillion-dimensional space, each dimensions having infinitely many possibilities, but you can map every single one of these points to a likelihood. It is then reasonable to fit your network using these parameters: given that the empirical data did occur, it is reasonable to assume that they should be likely under your model.
That doesn't mean that your parameters are likely ! Just that under these parameters, the observed value is likely. Statistical estimation is usually not a closed problem with a single solution (like solving an equation might be, and where you would have a probability of 1), but we need to find a best solution, according to some metric. Likelihood is such a metric, and is used widely because it has some interesting properties:
It makes intuitive sense
It's reasonably simple to compute, fit and optimise, for a large family of models
For normal variables (which tend to crop up everywhere) MLE gives the same results as other methods, such as least-squares estimations
Its formulation in terms of conditional probabilities makes it easy to use/manipulate it in Bayesian frameworks
For my class project, I am working on the Kaggle competition - Don't get kicked
The project is to classify test data as good/bad buy for cars. There are 34 features and the data is highly skewed. I made the following choices:
Since the data is highly skewed, out of 73,000 instances, 64,000 instances are bad buy and only 9,000 instances are good buy. Since building a decision tree would overfit the data, I chose to use kNN - K nearest neighbors.
After trying out kNN, I plan to try out Perceptron and SVM techniques, if kNN doesn't yield good results. Is my understanding about overfitting correct?
Since some features are numeric, I can directly use the Euclid distance as a measure, but there are other attributes which are categorical. To aptly use these features, I need to come up with my own distance measure. I read about Hamming distance, but I am still unclear on how to merge 2 distance measures so that each feature gets equal weight.
Is there a way to find a good approximate for value of k? I understand that this depends a lot on the use-case and varies per problem. But, if I am taking a simple vote from each neighbor, how much should I set the value of k? I'm currently trying out various values, such as 2,3,10 etc.
I researched around and found these links, but these are not specifically helpful -
a) Metric for nearest neighbor, which says that finding out your own distance measure is equivalent to 'kernelizing', but couldn't make much sense from it.
b) Distance independent approximation of kNN talks about R-trees, M-trees etc. which I believe don't apply to my case.
c) Finding nearest neighbors using Jaccard coeff
Please let me know if you need more information.
Since the data is unbalanced, you should either sample an equal number of good/bad (losing lots of "bad" records), or use an algorithm that can account for this. I think there's an SVM implementation in RapidMiner that does this.
You should use Cross-Validation to avoid overfitting. You might be using the term overfitting incorrectly here though.
You should normalize distances so that they have the same weight. By normalize I mean force to be between 0 and 1. To normalize something, subtract the minimum and divide by the range.
The way to find the optimal value of K is to try all possible values of K (while cross-validating) and chose the value of K with the highest accuracy. If a "good" value of K is fine, then you can use a genetic algorithm or similar to find it. Or you could try K in steps of say 5 or 10, see which K leads to good accuracy (say it's 55), then try steps of 1 near that "good value" (ie 50,51,52...) but this may not be optimal.
I'm looking at the exact same problem.
Regarding the choice of k, it's recommended be an odd value to avoid getting "tie votes".
I hope to expand this answer in the future.
I have a problem in that I need to implement an algorithm on an FPGA that requires a large array of data that is too large to fit into block or distributed memory. The array contains complex fixed-point values, and it turns out that I can do a good job by reducing the total number of stored values through decimation and then linearly interpolating the interim values on demand.
Though I have DSP blocks (and so fixed-point hardware multipliers) which could be used trivially for real and imaginary part interpolation, I actually want to do the interpolation on the amplitude and angle (of the polar form of the complex number) and then convert the result to real-imaginary form. The data can be stored in polar form if it improves things.
I think my question boils down to this: How should I quickly convert between polar complex numbers and real-imaginary complex numbers (and back again) on an FPGA (noting availability of DSP hardware)? The solution need not be exact, just close, but be speed optimised. Alternatively, better strategies are gladly received!
edit: I know about cordic techniques, so this would be how I would do it in the absence of a better idea. Are there refinements specific to this problem I could invoke?
Another edit: Following from #mbschenkel's question, and some more thinking on my part, I wanted to know if there were any known tricks specific to the problem of polar interpolation.
In my case, the dominant variation between samples is a phase rotation, with a slowly varying amplitude. Since the sampling grid is known ahead of time and is regular, one trick could be to precompute some complex interpolation factors. So, for two complex values a and b, if we wish to find (N-1) intermediate equally spaced values, we can precompute the factor
scale = (abs(b)/abs(a))**(1/N)*exp(1j*(angle(b)-angle(a)))/N)
and then find each intermediate value iteratively as val[n] = scale * val[n-1] where val[0] = a.
This works well for me as I need the samples in order and I compute them all. For small variations in amplitude (i.e. abs(b)/abs(a) ~= 1) and 0 < n < N, (abs(b)/abs(a))**(n/N) is approximately linear (though linear is not necessarily better).
The above is all very good, but still results in a complex multiplication. Are there other options for approximating this? I'm interested in resource and speed constraints, not accuracy. I know I can do the rotation with CORDIC, but still need a pair of multiplications for the scaling, so I'm adding lots of complexity and resource usage for potentially limited results. I don't really have a feel for the convergence of CORDIC, so perhaps I just truncate early, or use lots of resources to converge quickly.
I know that principal component analysis does a SVD on a matrix and then generates an eigen value matrix. To select the principal components we have to take only the first few eigen values. Now, how do we decide on the number of eigen values that we should take from the eigen value matrix?
To decide how many eigenvalues/eigenvectors to keep, you should consider your reason for doing PCA in the first place. Are you doing it for reducing storage requirements, to reduce dimensionality for a classification algorithm, or for some other reason? If you don't have any strict constraints, I recommend plotting the cumulative sum of eigenvalues (assuming they are in descending order). If you divide each value by the total sum of eigenvalues prior to plotting, then your plot will show the fraction of total variance retained vs. number of eigenvalues. The plot will then provide a good indication of when you hit the point of diminishing returns (i.e., little variance is gained by retaining additional eigenvalues).
There is no correct answer, it is somewhere between 1 and n.
Think of a principal component as a street in a town you have never visited before. How many streets should you take to get to know the town?
Well, you should obviously visit the main street (the first component), and maybe some of the other big streets too. Do you need to visit every street to know the town well enough? Probably not.
To know the town perfectly, you should visit all of the streets. But what if you could visit, say 10 out of the 50 streets, and have a 95% understanding of the town? Is that good enough?
Basically, you should select enough components to explain enough of the variance that you are comfortable with.
As others said, it doesn't hurt to plot the explained variance.
If you use PCA as a preprocessing step for a supervised learning task, you should cross validate the whole data processing pipeline and treat the number of PCA dimension as an hyperparameter to select using a grid search on the final supervised score (e.g. F1 score for classification or RMSE for regression).
If cross-validated grid search on the whole dataset is too costly try on a 2 sub samples, e.g. one with 1% of the data and the second with 10% and see if you come up with the same optimal value for the PCA dimensions.
There are a number of heuristics use for that.
E.g. taking the first k eigenvectors that capture at least 85% of the total variance.
However, for high dimensionality, these heuristics usually are not very good.
Depending on your situation, it may be interesting to define the maximal allowed relative error by projecting your data on ndim dimensions.
Matlab example
I will illustrate this with a small matlab example. Just skip the code if you are not interested in it.
I will first generate a random matrix of n samples (rows) and p features containing exactly 100 non zero principal components.
n = 200;
p = 119;
data = zeros(n, p);
for i = 1:100
data = data + rand(n, 1)*rand(1, p);
end
The image will look similar to:
For this sample image, one can calculate the relative error made by projecting your input data to ndim dimensions as follows:
[coeff,score] = pca(data,'Economy',true);
relativeError = zeros(p, 1);
for ndim=1:p
reconstructed = repmat(mean(data,1),n,1) + score(:,1:ndim)*coeff(:,1:ndim)';
residuals = data - reconstructed;
relativeError(ndim) = max(max(residuals./data));
end
Plotting the relative error in function of the number of dimensions (principal components) results in the following graph:
Based on this graph, you can decide how many principal components you need to take into account. In this theoretical image taking 100 components result in an exact image representation. So, taking more than 100 elements is useless. If you want for example maximum 5% error, you should take about 40 principal components.
Disclaimer: The obtained values are only valid for my artificial data. So, do not use the proposed values blindly in your situation, but perform the same analysis and make a trade off between the error you make and the number of components you need.
Code reference
Iterative algorithm is based on the source code of pcares
A StackOverflow post about pcares
I highly recommend the following paper by Gavish and Donoho: The Optimal Hard Threshold for Singular Values is 4/sqrt(3).
I posted a longer summary of this on CrossValidated (stats.stackexchange.com). Briefly, they obtain an optimal procedure in the limit of very large matrices. The procedure is very simple, does not require any hand-tuned parameters, and seems to work very well in practice.
They have a nice code supplement here: https://purl.stanford.edu/vg705qn9070