What is the meaning of words to keep attribute in Weka StringToWord filter. Is it better to have higher value or not, for getting real results?
In general, it is a good idea to set the limit as high as possible in order to retain as many words as possible. Words with small frequencies can marginally help the classifiers you induce later.
Keeping too many words may look like a bad idea for a matter of efficiency - the higher the number of attributes, the longer it will take to learn the model. However, you can filter the words to keep the most predictive ones using the AttributeSelection filter with the Ranker function and the InfoGainAttributeEval measure. In fact, you can play with the theshold in the AttrivuteSelection filter in order to keep a relatively small number of very predictive words, with independence of their relative frequency.
Additionally, do not forget to set the flag doNotOperatePerClassBasis to true in order to keep all the words relevant to all classes.
Related
Consider a search system where the user submits a query ‘query’ and retrieves products based on some ranking algorithm. Assume that these products are ordered according to their quality such that p_0, p_1, …, p_10 and so on.
I would like to generate vector embeddings that mimic this ranking algorithm. The closest product vector to a query vector should ideally be p_0, the next one should be p_1 and so on.
I have tried to building word2vec embeddings for products by feeding products that have appeared in the same search session as sentences. Then, I have calculated the weighted average of product vectors to find query vectors to make the query vector closer to the top result. Although the closest result is usually the best result for a given query, the subsequent results include some results that would never appear as a top result.
Is there a trick that the word2vec can learn the ranking algorithm or any other techniques that I can try? I have looked into multi-dimensional vector scaling with non-metric distances but it did not seem scalable to me for more than 100Ks of products.
There's no one trick – just iteratively improving your representations, & training set, & ranking methods to better meet your goals.
Word2vec-based representations can often help, but are still fairly simple & centered on individual words – whose senses may vary based on context & position in ways that a simple weighted-average-of-tokens fails to capture.
You may want to represent 'products' by more than just a string-of-word-tokens – to include other properties, as well. These could be scalar values like prices or various other kinds of ratings/properties, or extra synthetic labels, such as the result of other salient groupings (whether hand-edited or learned).
And even if just working with natural-language product descriptions – like product names, or descriptions, or reviews – there are other more-sophisticated text-representations that can be trained or used – such as sentence/document embeddings using deeper-networks than plain word2vec.
Most generically, if you have a bunch of quantitative representations of candidate results, and a query, and want to use some initial examples of "good" results to bootstrap more generalizable rules for scoring top results, you are attempting a "learning-to-rank" process:
https://en.wikipedia.org/wiki/Learning_to_rank
To suggest more specific steps would require a more specific description of inputs/outputs/goals, & what's been tried, and how what's been tried has failed.
For example, are your queries always just textual product names? In such a case, maybe plain keyword search is the central technology required – with things like word-vector-modelling just a tweak for handling some tough cases, like expanding the results, or adding more contrast to the rankings, when results are too few or to many.
Or, can you detect key gaps in the modeling related to exactly those cases where "results include some results that would [ideally] never appear as a top result"? If certain things like rare (poorly-modeled) words, or important qualities not yet captured in the model, seem to be to blame for such cases, that will guide the potential set of corrective changes.
I am using H2O machine learning package to do natural language predictions, including the functions h2o.word2vec and h2o.transform. I need sentence level aggregation, which is provided by the AVERAGE parameter value:
h2o.transform(word2vec, words, aggregate_method = c("NONE", "AVERAGE"))
However, in my case I strongly wish to avoid equal weighting of "the" and "platypus" for example.
Here's a scheme I concocted to achieve custom word-weightings. If H2O's word2vec "AVERAGE" option uses all the words including duplicates that might appear, then I could effect a custom word weighting when calling h2o.transform by adding additional duplicates of certain words to my sentences, when I want to weight them more heavily than other words.
Can any H2O experts confirm that that the word2vec AVERAGE parameter is using all the words rather than just the unique words when computing AVERAGE of the words in sentence?
Alternatively, is there a better way? I tried but I find myself unable to imagine any correct math to multiply the sentence average by some factor, after it was already computed.
Yes, h2o.transform will consider each occurrence of a word for the averaging, not just the unique words. Your trick will therefore work.
There is currently no direct way to provide user defined weights. You could probably do an ugly hack and weight directly the word embeddings but that won't be a straightforward solution I could recommend.
We can add this feature to H2O. I would love to hear what API would work for you (how would you like to provide the weights).
I am new in rapid miner 5, just want to know how to find noise in my data and show them in chart and how to delete them?
A complex problem because it depends what you mean by noise.
If you mean finding individual attributes whose values are plain wrong then you could plot a histogram view and work out some sort of limits on what constitutes a valid value. You could then impose that rule by using Filter Examples to remove them.
If you mean finding attributes that have some sort of random jitter applied to them it would be difficult to detect these. Only by knowing beforehand what the expected shape of the distribution is could you compare with observation and do something about it. However, the action to take is by no means obvious.
If you mean finding examples within an example set that are obviously different from other examples then you could consider using the various outlier functions. The simplest one to get started is Detect Outlier (Distances). This finds a set number of outliers (default 10) based on a distance calculation that uses all the attributes for examples. It creates a new attribute called outlier that is set to true or false. You could then use the Filter Examples operator to remove those that are set to true.
Hope that helps at least as a start.
I am relatively new to the data mining area and have been experimenting with Weka.
I have a dataset which consists of almost 8000 records related to customers and items they have purchased. 58% of this data set has missing values for the "Gender" attribute.
I want to find the missing gender values based on the other data I do have.
I first thought I could do this using a classifier algorithm in Weka using a training set to build a model. Based on examples I saw online, I tried this with pretty much all the available algorithms available in Weka using a training set that consisted of 60-80% of the data which did not have missing values. This gave me a lower accuracy rate than I wanted (80-86% depending on the algorithm used)
Did I go about this correctly? Is there a way to improve this accuracy? I experimented with using different attributes, different pre-processing of the data etc.
I also tried using the ReplaceMissingValues filter on the complete dataset to see how that would handle the missing values. However, it just changed all the missing values to "Female" which obviously cannot be the case. So I'm wondering also wondering if I need to use this filter in my situation or not.
It sounds like you went about it in the correct way. The ReplaceMissingValues filter replaces the missing values with the most frequent of the non-missing values I think, so it is not what you want in this case.
A better way to get an idea of the true accuracy of your gender-predictor would be to use cross-validation instead of the training/test split (Weka has a separate option for that). 80-86% may seem low, but keep in mind that random guessing will only get you about 50%, so it's still a lot better than that. To try to get better performance, pick a classifier that performs well and then play with its parameters until you get better performance. This is likely to be quite labour-intensive (although you could of course use automated methods for tuning, see e.g. Auto-WEKA), but the only way to improve the performance.
You can also combine the algorithm you choose with a separate feature selection step (Weka has a special meta-classifier for this). This may improve performance, but again you'll have to experiment to find the particular configuration that works for you.
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.