I wanted to know whether I can use the predict option for ancillary parameters (maximum likelihood ) program as follows (I estimated lnsigma and so sigma is the ancillary parameter in the model):
predict lnsigma, eq(lnsigma)
gen sigma=exp(lnsigma)
I also would like to know whether we can use above for heteroscedastic model.
Thank you in advance.
That sounds correct. I would be more explicit by typing predict lnsigma, xb eq(lnsigma). This way your code will not be broken when someone later on desides to write a prediction program for your estimation program and sets the default to something different than the linear prediction.
You can also do it in one line:
predictnl sigma = exp(xb(#2))
This assumes that lnsigma is the second equation in your model. If it is the third equation you replace xb(#2) with xb(#3). predictnl is also also an easy way of using the delta method to predict standard errors and confidence intervals for sigma.
I assume this is your own Stata program. If that is true, then you also have a third option: You can create your own prediction program, which Stata's predict command will recongnize. You can find some useful tricks on how to do that here: http://www.stata.com/help.cgi?_pred_se
Related
I built a pymc3 model using the DensityDist distribution. I have four parameters out of which 3 use Metropolis and one uses NUTS (this is automatically chosen by the pymc3). However, I get two different UserWarnings
1.Chain 0 contains number of diverging samples after tuning. If increasing target_accept does not help try to reparameterize.
MAy I know what does reparameterize here mean?
2. The acceptance probability in chain 0 does not match the target. It is , but should be close to 0.8. Try to increase the number of tuning steps.
Digging through a few examples I used 'random_seed', 'discard_tuned_samples', 'step = pm.NUTS(target_accept=0.95)' and so on and got rid of these user warnings. But I couldn't find details of how these parameter values are being decided. I am sure this might have been discussed in various context but I am unable to find solid documentation for this. I was doing a trial and error method as below.
with patten_study:
#SEED = 61290425 #51290425
step = pm.NUTS(target_accept=0.95)
trace = sample(step = step)#4000,tune = 10000,step =step,discard_tuned_samples=False)#,random_seed=SEED)
I need to run these on different datasets. Hence I am struggling to fix these parameter values for each dataset I am using. Is there any way where I give these values or find the outcome (if there are any user warnings and then try other values) and run it in a loop?
Pardon me if I am asking something stupid!
In this context, re-parametrization basically is finding a different but equivalent model that it is easier to compute. There are many things you can do depending on the details of your model:
Instead of using a Uniform distribution you can use a Normal distribution with a large variance.
Changing from a centered-hierarchical model to a
non-centered
one.
Replacing a Gaussian with a Student-T
Model a discrete variable as a continuous
Marginalize variables like in this example
whether these changes make sense or not is something that you should decide, based on your knowledge of the model and problem.
Is there a way to incorporate the uncertainties on my data set into the result of the Savitzky Golay fit? Since I am not passing this information into the function, I asume that it is simply calcuating the 'best fit' via an unweighted least-squares process. I am currently working with data that has non-uniform uncertainty, and so the fit of the data could be improved by including the errors that I have for my main dataset.
The wikipedia page for the Savitzky-Golay filter suggests how I might go about alter the process of calculating the coefficients of the fit, and I am staring at the code for scipy.signal.savgol_filter, but I cannot get my head around what I need to adjust so that this will do what I want it to.
Are there any ready-made weighted SG filters floating about? I find it hard to believe that no-one else has ever needed this tool in Python, but maybe I have missed something.
Check out this Python module: https://github.com/surhudm/savitzky_golay_with_errors
This python script improves upon the traditional Savitzky-Golay filter
by accounting for errors or covariance in the data. The inputs and
arguments are all modelled after scipy.signal.savgol_filter
Matlab function sgolayfilt supports weights. Check the documentation.
I am trying to predict whether a particular service ticket raised by client needs a code change.
I have training data.
I have around 17k data points with problem description and tag (Y for code change required and N for no code change)
I did TF-IDF and it gave me 27k features. So I tried to fit RandomForestClassifier (sklearn python) with this 17k x 27k matrix.
I am getting very low scores on test set while training accuracy is very high.
Precision on train set: 89%
Precision on test set: 21%
Can someone suggest any workarounds?
I am using this model now:
sklearn.RandomForestClassifier(n_jobs=3,n_estimators=100,class_weight='balanced',max_features=None,oob_score=True)
Please help!
EDIT:
I have 11k training data with 900 positives (skewed). I tried LinearSVC sparsify but didn't work as well as Truncated SVD (Latent Semantic Indexing). maxFeatures=None performs better on the test set than without it.
I have also tried SVM, logistic (l2 and l1), ExtraTrees. RandonForest still is working best.
Right now, going at 92% precision on positives but recall is 3% only
Any other suggestions would be appreciated!
Update:
Feature engineering helped a lot. I pulled features out of the air (len of chars, len of words, their, difference, ratio, day of week the problem was of reported, day of month, etc) and now I am at 19-20% recall with >95% accuracy.
Food for your thoughts on using word2vec average vectors as deep features for the free text instead of tf-idf or bag of words ???
[edited]
Random forest handles more features than data points quite fine. RF is e.g. used for micro-array studies with e.g. a 100:5000 data point/feature ratio or in single-nucleotide_polymorphism(SNP) studies with e.g 5000:500,000 ratio.
I do disagree with the diagnose provided by #ncfirth, but the suggested treatment of variable selection may help anyway.
Your default random forest is not badly overfitted. It is just not meaningful to pay any attention to a non-cross validated training set prediction performance for a RF model, because any sample will end in the terminal nodes/leafs it has itself defined. But the overall ensemble model is still robust.
[edit] If you would change the max_depth or min_samples_split, the training precision would probably drop, but that is not the point. The non-cross validated training error/precision of a random forest model or many other ensemble models simply does not estimate anything useful.
[I did before edit confuse max_features with n_estimators, sry I mostly use R]
Setting max_features="none" is not random forest, but rather 'bagged trees'. You may benefit from a somewhat lower max_features which improve regularization and speed, maybe not. I would try lowering max_features to somewhere between 27000/3 and sqrt(27000), the typical optimal range.
You may achieve better test set prediction performance by feature selection. You can run one RF model, keep the top ~5-50% most important features and then re-run the model with fewer features. "L1 lasso" variable selection as ncfirth suggests may also be a viable solution.
Your metric of prediction performance, precision, may not be optimal in case unbalanced data or if the cost of false-negative and false-positive is quite different.
If your test set is still predicted much worse than the out-of-bag cross-validated training set, you may have problems with your I.I.D. assumptions that any supervised ML model rely on or you may need to wrap the entire data processing in an outer cross-validation loop, to avoid over optimistic estimation of prediction performance due to e.g. the variable selection step.
Seems like you've overfit on your training set. Basically the model has learnt noise on the data rather than the signal. There are a few ways to combat this, but it seems fairly obvious that you're model has overfit because of the incredibly large number of features you're feeding it.
EDIT:
It seems I was perhaps too quick to jump to the conclusion of overfitting, however this may still be the case (left as an exercise to the reader!). However feature selection may still improve the generalisability and reliability of your model.
A good place to start for removing features in scikit-learn would be here. Using sparsity is a fairly common way to perform feature selection:
from sklearn.svm import LinearSVC
from sklearn.feature_selection import SelectFromModel
import numpy as np
# Create some data
X = np.random.random((1800, 2700))
# Boolean labels as the y vector
y = np.random.random(1800)
y = y > 0.5
y = y.astype(bool)
lsvc = LinearSVC(C=0.05, penalty="l1", dual=False).fit(X, y)
model = SelectFromModel(lsvc, prefit=True)
X_new = model.transform(X)
print X_new.shape
Which returns a new matrix of shape (1800, 640). You can tune the number of features selected by altering the C parameter (called the penalty parameter in scikit-learn but sometimes called the sparsity parameter).
I am trying to estimate a maximum likelihood model and it is running into convergence problems in Stata. The actual model is quite complicated, but it converges with no troubles in R when it is supplied with appropriate starting values. I however cannot seem to get Stata to accept the starting values I provide.
I have included a simple example below estimating the mean of a poisson distribution. This is not the actual model I am trying to estimate, but it demonstrates my problem. I set the trace variable, which allows you to see the parameters as Stata searches the likelihood surface.
Although I use init to set a starting value of 0.5, the first iteration still shows that Stata is trying a coefficient of 4.
Why is this? How can I force the estimation procedure to use my starting values?
Thanks!
generate y = rpoisson(4)
capture program drop mypoisson
program define mypoisson
args lnf mu
quietly replace `lnf' = $ML_y1*ln(`mu') - `mu' - lnfactorial($ML_y1)
end
ml model lf mypoisson (mean:y=)
ml init 0.5, copy
ml maximize, iterations(2) trace
Output:
Iteration 0:
Parameter vector:
mean:
_cons
r1 4
Added: Stata doesn't ignore the initial value. If you look at the output of the ml maximize command, the first line in the listing will be titled
initial: log likelihood =
Following the equal sign is the value of the likelihood for the parameter value set in the init statement.
I don't know how the search(off) or search(norescale) solutions affect the subsequent likelihood calculations, so these solution might still be worthwhile.
Original "solutions":
To force a start at your initial value, add the search(off) option to ml maximize:
ml maximize, iterate(2) trace search(off)
You can also force a use of the initial value with search(norescale). See Jeff Pitblado's post at http://www.stata.com/statalist/archive/2006-07/msg00499.html.
I'm using Weka and would like to perform regression with random forests. Specifically, I have a dataset:
Feature1,Feature2,...,FeatureN,Class
1.0,X,...,1.4,Good
1.2,Y,...,1.5,Good
1.2,F,...,1.6,Bad
1.1,R,...,1.5,Great
0.9,J,...,1.1,Horrible
0.5,K,...,1.5,Terrific
.
.
.
Rather than learning to predict the most likely class, I want to learn the probability distribution over the classes for a given feature vector. My intuition is that using just the RandomForest model in Weka would not be appropriate, since it would be attempting to minimize its absolute error (maximum likelihood) rather than its squared error (conditional probability distribution). Is that intuition right? Is there a better model to be using if I want to perform regression rather than classification?
Edit: I'm actually thinking now that in fact it may not be a problem. Presumably, classifiers are learning the conditional probability P(Class | Feature1,...,FeatureN) and the resulting classification is just finding the c in Class that maximizes that probability distribution. Therefore, a RandomForest classifier should be able to give me the conditional probability distribution. I just had to think about it some more. If that's wrong, please correct me.
If you want to predict the probabilities for each class explicitly, you need different input data. That is, you would need to replace the value to predict. Instead of one data set with the class label, you would need n data sets (for n different labels) with aggregated data for each unique feature vector. Your data would look something like
Feature1,...,Good
1.0,...,0.5
0.3,...,1.0
and
Feature1,...,Bad
1.0,...,0.8
0.3,...,0.1
and so on. You would need to learn one model for each class and run them separately on any data to be classified. That is, for each label you learn a model to predict a number that is the probability of being in that class, given a feature vector.
If you don't need the probabilities to be predicted explicitly, have a look at the Bayesian classifiers in Weka, which make use of probabilities in the models that they learn.