What does "Naive" Bayes mean in Machine learning?
Naive Bayes in machine learning typically refers to a set of supervised learning algorithms that apply the Bayes' theorem. It's essentially a "classifier" that helps you classify things based on a series of independent "naive" assumptions. For example if you wanted to use machine learning to help you identify potential fruit... taking a banana, its curved, yellow, and may be 10 inches long. Each of those properties, 'curved', 'yellow', '10 inches long' are all independent properties that come together to form a 'probability' that a fruit is a banana. With this 'naive' bayes classifier, in the future if there are other kinds of 'fruit' or different images, descriptions of fruit that have similar properties, using machine learning, your naive bayes classifier can classify those future fruits or unknown things as bananas correctly... (or incorrectly) in which you'll probably want to identify more 'naive' features to make your classifier more accurate like for example, there might be a 'blackened tip' or 'have a slight greenish color'.
It is called naïve because the model assumes independence between the features. This is a strong assumption which usually is not correct and that is the reason of the name.
Nevertheless, naïve Bayes is quite efficient and in practice is known for giving good results.
A naive Bayes classifier is an algorithm that uses Bayes' theorem to classify objects. Naive Bayes classifiers assume strong, or naive, independence between attributes of data points. Popular uses of naive Bayes classifiers include spam filters, text analysis and medical diagnosis.
What makes a naive Bayes classifier naive is its assumption that all attributes of a data point under consideration are independent of each other. A classifier sorting fruits into apples and oranges would know that apples are red, round and are a certain size, but would not assume all these things at once. Oranges are round too, after all.
A naive Bayes classifier is not a single algorithm, but a family of machine learning algorithms that make uses of statistical independence. These algorithms are relatively easy to write and run more efficiently than more complex Bayes algorithms.
The most popular application is spam filters. A spam filter looks at email messages for certain key words and puts them in a spam folder if they match.
Despite the name, the more data it gets, the more accurate a naive Bayes classifier becomes, such as from a user flagging email messages in an inbox for spam.
Related
Actually I am working on Naive Bayes classifier for filtering mails. I have achieved accuracy of 95% in SPAM detection and 94 % in HAM detection, but I believe it can be further improved with association rule mining. I am calculating likelihood and prior probabilities of the words in mails from training data set, and mapping the testing mail to either of SPAM or HAM class as given below,
where,
p(d/c) denotes probability of document d being in class c.
p(c) denotes probability of particular class(SPAM or HAM in my case).
p(f1,f2,f3...fn/c) denotes likelihood of words f1,f2...fn being in class c.
but while arriving at equation no. 2.7, we assume bag of words assumption and condition independence, which approximates accuracy( Which is assumed for sake of easiness).
e.g. Likeliness of word lottery being in SPAM mail with presence of word lucky should be greater than the same with presence of word my_name(mahesh). So presence of words and their position do affect probabilities.
Therefore there should be some associative model in accordance with Naive Bayes to further improve accuracy.
If I can rephrase your question like this:
"Will relaxing the conditional independence assumption of Naive Bayes improve my classifier's performance?"
Then the answer is a surprising and counterintuitive "No."
Generally speaking, Naive Bayes classifiers, which impose strict class-conditional independence between features, will offer same-or-better performance than more general Bayesian networks, which allow for richer dependencies (and that dependence structure can even be learned from data, although generally not exactly).
The reason is that, while Naive Bayes will generally get the probabilities wrong, it will generally get the decision boundary right [1].
So: you are probably better off just making the bag-of-words assumption.
[1] http://web.cs.ucdavis.edu/~vemuri/classes/ecs271/Bayesian.pdf
I'm working on a problem that involves classifying a large database of texts. The texts are very short (think 3-8 words each) and there are 10-12 categories into which I wish to sort them. For the features, I'm simply using the tf–idf frequency of each word. Thus, the number of features is roughly equal to the number of words that appear overall in the texts (I'm removing stop words and some others).
In trying to come up with a model to use, I've had the following two ideas:
Naive Bayes (likely the sklearn multinomial Naive Bayes implementation)
Support vector machine (with stochastic gradient descent used in training, also an sklearn implementation)
I have built both models, and am currently comparing the results.
What are the theoretical pros and cons to each model? Why might one of these be better for this type of problem? I'm new to machine learning, so what I'd like to understand is why one might do better.
Many thanks!
The biggest difference between the models you're building from a "features" point of view is that Naive Bayes treats them as independent, whereas SVM looks at the interactions between them to a certain degree, as long as you're using a non-linear kernel (Gaussian, rbf, poly etc.). So if you have interactions, and, given your problem, you most likely do, an SVM will be better at capturing those, hence better at the classification task you want.
The consensus for ML researchers and practitioners is that in almost all cases, the SVM is better than the Naive Bayes.
From a theoretical point of view, it is a little bit hard to compare the two methods. One is probabilistic in nature, while the second one is geometric. However, it's quite easy to come up with a function where one has dependencies between variables which are not captured by Naive Bayes (y(a,b) = ab), so we know it isn't an universal approximator. SVMs with the proper choice of Kernel are (as are 2/3 layer neural networks) though, so from that point of view, the theory matches the practice.
But in the end it comes down to performance on your problem - you basically want to choose the simplest method which will give good enough results for your problem and have a good enough performance. Spam detection has been famously solvable by just Naive Bayes, for example. Face recognition in images by a similar method enhanced with boosting etc.
Support Vector Machine (SVM) is better at full-length content.
Multinomial Naive Bayes (MNB) is better at snippets.
MNB is stronger for snippets than for longer documents. While (Ng and Jordan,
2002) showed that NB is better than SVM/logistic
regression (LR) with few training cases, MNB is also better with short documents. SVM usually beats NB when it has more than 30–50 training cases, we show that MNB is still better on snippets even with relatively large training sets (9k cases).
Inshort, NBSVM seems to be an appropriate and very strong baseline for sophisticated classification text data.
Source Code: https://github.com/prakhar-agarwal/Naive-Bayes-SVM
Reference: http://nlp.stanford.edu/pubs/sidaw12_simple_sentiment.pdf
Cite: Wang, Sida, and Christopher D. Manning. "Baselines and bigrams:
Simple, good sentiment and topic classification." Proceedings of the
50th Annual Meeting of the Association for Computational Linguistics:
Short Papers-Volume 2. Association for Computational Linguistics,
2012.
I am new to Machine Learning.I am working on a project where the machine learning concept need to be applied.
Problem Statement:
I have large number(say 3000)key words.These need to be classified into seven fixed categories.Each category is having training data(sample keywords).I need to come with a algorithm, when a new keyword is passed to that,it should predict to which category this key word belongs to.
I am not aware of which text classification technique need to applied for this.do we have any tools that can be used.
Please help.
Thanks in advance.
This comes under linear classification. You can use naive-bayes classifier for this. Most of the ml frameworks will have an implementation for naive-bayes. ex: mahout
Yes, I would also suggest to use Naive Bayes, which is more or less the baseline classification algorithm here. On the other hand, there are obviously many other algorithms. Random forests and Support Vector Machines come to mind. See http://machinelearningmastery.com/use-random-forest-testing-179-classifiers-121-datasets/ If you use a standard toolkit, such as Weka, Rapidminer, etc. these algorithms should be available. There is also OpenNLP for Java, which comes with a maximum entropy classifier.
You could use the Word2Vec Word Cosine distance between descriptions of each your category and keywords in the dataset and then simple match each keyword to a category with the closest distance
Alternatively, you could create a training dataset from already matched to category, keywords and use any ML classifier, for example, based on artificial neural networks by using vectors of keywords Cosine distances to each category as an input to your model. But it could require a big quantity of data for training to reach good accuracy. For example, the MNIST dataset contains 70000 of the samples and it allowed me reach 99,62% model's cross validation accuracy with a simple CNN, for another dataset with only 2000 samples I was able reached only about 90% accuracy
There are many classification algorithms. Your example looks to be a text classification problems - some good classifiers to try out would be SVM and naive bayes. For SVM, liblinear and libshorttext classifiers are good options (and have been used in many industrial applcitions):
liblinear: https://www.csie.ntu.edu.tw/~cjlin/liblinear/
libshorttext:https://www.csie.ntu.edu.tw/~cjlin/libshorttext/
They are also included with ML tools such as scikit-learna and WEKA.
With classifiers, it is still some operation to build and validate a pratically useful classifier. One of the challenges is to mix
discrete (boolean and enumerable)
and continuous ('numbers')
predictive variables seamlessly. Some algorithmic preprocessing is generally necessary.
Neural networks do offer the possibility of using both types of variables. However, they require skilled data scientists to yield good results. A straight-forward option is to use an online classifier web service like Insight Classifiers to build and validate a classifier in one go. N-fold cross validation is being used there.
You can represent the presence or absence of each word in a separate column. The outcome variable is desired category.
I have been trying to figure out the correlation between the error rate and the number of features in both of these models. I watched some videos, and the creator of the video said that a simple model can be better than a complicated model. So I figured that the more features I had the greater the error rate would be. This did not prove to be true in my work, and when I had less features the error rate went up. I'm not sure if I'm doing this incorrectly, or if the guy in the video made a mistake. Can someone care to explain? I also am curious how features relate to Logistic Regression's error rate as well.
Naive Bayes and Logistic Regression are a "generative-discriminative pair," meaning they have the same model form (a linear classifier), but they estimate parameters in different ways.
For feature x and label y, naive Bayes estimates a joint probability p(x,y) = p(y)*p(x|y) from the training data (that is, builds a model that could "generate" the data), and uses Bayes Rule to predict p(y|x) for new test instances. On the other hand, logistic regression estimates p(y|x) directly from the training data by minimizing an error function (which is more "discrimative").
These differences have implications for error rate:
When there are very few training instances, logistic regression might "overfit," because there isn't enough data to estimate p(y|x) reliably. Naive Bayes might do better because it models the entire joint distribution.
When the feature set is large (and sparse, like word features in text classification) naive Bayes might "double count" features that are correlated with each other, because it assumes that each p(x|y) event is independent, when they are not. Logistic regression can do a better job by naturally "splitting the difference" among these correlated features.
If the features really are (mostly) conditionally independent, both models might actually improve with more and more features, provided there are enough data instances. The problem comes when the training set size is small relative to the number of features. Priors on naive Bayes feature parameters, or regularization methods (like L1/Lasso or L2/Ridge) on logistic regression can help in these cases.
I know SVMs are supposedly 'ANN killers' in that they automatically select representation complexity and find a global optimum (see here for some SVM praising quotes).
But here is where I'm unclear -- do all of these claims of superiority hold for just the case of a 2 class decision problem or do they go further? (I assume they hold for non-linearly separable classes or else no-one would care)
So a sample of some of the cases I'd like to be cleared up:
Are SVMs better than ANNs with many classes?
in an online setting?
What about in a semi-supervised case like reinforcement learning?
Is there a better unsupervised version of SVMs?
I don't expect someone to answer all of these lil' subquestions, but rather to give some general bounds for when SVMs are better than the common ANN equivalents (e.g. FFBP, recurrent BP, Boltzmann machines, SOMs, etc.) in practice, and preferably, in theory as well.
Are SVMs better than ANN with many classes? You are probably referring to the fact that SVMs are in essence, either either one-class or two-class classifiers. Indeed they are and there's no way to modify a SVM algorithm to classify more than two classes.
The fundamental feature of a SVM is the separating maximum-margin hyperplane whose position is determined by maximizing its distance from the support vectors. And yet SVMs are routinely used for multi-class classification, which is accomplished with a processing wrapper around multiple SVM classifiers that work in a "one against many" pattern--i.e., the training data is shown to the first SVM which classifies those instances as "Class I" or "not Class I". The data in the second class, is then shown to a second SVM which classifies this data as "Class II" or "not Class II", and so on. In practice, this works quite well. So as you would expect, the superior resolution of SVMs compared to other classifiers is not limited to two-class data.
As far as i can tell, the studies reported in the literature confirm this, e.g., In the provocatively titled paper Sex with Support Vector Machines substantially better resolution for sex identification (Male/Female) in 12-square pixel images, was reported for SVM compared with that of a group of traditional linear classifiers; SVM also outperformed RBF NN, as well as large ensemble RBF NN). But there seem to be plenty of similar evidence for the superior performance of SVM in multi-class problems: e.g., SVM outperformed NN in protein-fold recognition, and in time-series forecasting.
My impression from reading this literature over the past decade or so, is that the majority of the carefully designed studies--by persons skilled at configuring and using both techniques, and using data sufficiently resistant to classification to provoke some meaningful difference in resolution--report the superior performance of SVM relative to NN. But as your Question suggests, that performance delta seems to be, to a degree, domain specific.
For instance, NN outperformed SVM in a comparative study of author identification from texts in Arabic script; In a study comparing credit rating prediction, there was no discernible difference in resolution by the two classifiers; a similar result was reported in a study of high-energy particle classification.
I have read, from more than one source in the academic literature, that SVM outperforms NN as the size of the training data decreases.
Finally, the extent to which one can generalize from the results of these comparative studies is probably quite limited. For instance, in one study comparing the accuracy of SVM and NN in time series forecasting, the investigators reported that SVM did indeed outperform a conventional (back-propagating over layered nodes) NN but performance of the SVM was about the same as that of an RBF (radial basis function) NN.
[Are SVMs better than ANN] In an Online setting? SVMs are not used in an online setting (i.e., incremental training). The essence of SVMs is the separating hyperplane whose position is determined by a small number of support vectors. So even a single additional data point could in principle significantly influence the position of this hyperplane.
What about in a semi-supervised case like reinforcement learning? Until the OP's comment to this answer, i was not aware of either Neural Networks or SVMs used in this way--but they are.
The most widely used- semi-supervised variant of SVM is named Transductive SVM (TSVM), first mentioned by Vladimir Vapnick (the same guy who discovered/invented conventional SVM). I know almost nothing about this technique other than what's it is called and that is follows the principles of transduction (roughly lateral reasoning--i.e., reasoning from training data to test data). Apparently TSV is a preferred technique in the field of text classification.
Is there a better unsupervised version of SVMs? I don't believe SVMs are suitable for unsupervised learning. Separation is based on the position of the maximum-margin hyperplane determined by support vectors. This could easily be my own limited understanding, but i don't see how that would happen if those support vectors were unlabeled (i.e., if you didn't know before-hand what you were trying to separate). One crucial use case of unsupervised algorithms is when you don't have labeled data or you do and it's badly unbalanced. E.g., online fraud; here you might have in your training data, only a few data points labeled as "fraudulent accounts" (and usually with questionable accuracy) versus the remaining >99% labeled "not fraud." In this scenario, a one-class classifier, a typical configuration for SVMs, is the a good option. In particular, the training data consists of instances labeled "not fraud" and "unk" (or some other label to indicate they are not in the class)--in other words, "inside the decision boundary" and "outside the decision boundary."
I wanted to conclude by mentioning that, 20 years after their "discovery", the SVM is a firmly entrenched member in the ML library. And indeed, the consistently superior resolution compared with other state-of-the-art classifiers is well documented.
Their pedigree is both a function of their superior performance documented in numerous rigorously controlled studies as well as their conceptual elegance. W/r/t the latter point, consider that multi-layer perceptrons (MLP), though they are often excellent classifiers, are driven by a numerical optimization routine, which in practice rarely finds the global minimum; moreover, that solution has no conceptual significance. On the other hand, the numerical optimization at the heart of building an SVM classifier does in fact find the global minimum. What's more that solution is the actual decision boundary.
Still, i think SVM reputation has declined a little during the past few years.
The primary reason i suspect is the NetFlix competition. NetFlix emphasized the resolving power of fundamental techniques of matrix decomposition and even more significantly t*he power of combining classifiers. People combined classifiers long before NetFlix, but more as a contingent technique than as an attribute of classifier design. Moreover, many of the techniques for combining classifiers are extraordinarily simple to understand and also to implement. By contrast, SVMs are not only very difficult to code (in my opinion, by far the most difficult ML algorithm to implement in code) but also difficult to configure and implement as a pre-compiled library--e.g., a kernel must be selected, the results are very sensitive to how the data is re-scaled/normalized, etc.
I loved Doug's answer. I would like to add two comments.
1) Vladimir Vapnick also co-invented the VC dimension which is important in learning theory.
2) I think that SVMs were the best overall classifiers from 2000 to 2009, but after 2009, I am not sure. I think that neural nets have improved very significantly recently due to the work in Deep Learning and Sparse Denoising Auto-Encoders. I thought I saw a number of benchmarks where they outperformed SVMs. See, for example, slide 31 of
http://deeplearningworkshopnips2010.files.wordpress.com/2010/09/nips10-workshop-tutorial-final.pdf
A few of my friends have been using the sparse auto encoder technique. The neural nets build with that technique significantly outperformed the older back propagation neural networks. I will try to post some experimental results at artent.net if I get some time.
I'd expect SVM's to be better when you have good features to start with. IE, your features succinctly capture all the necessary information. You can see if your features are good if instances of the same class "clump together" in the feature space. Then SVM with Euclidian kernel should do the trick. Essentially you can view SVM as a supercharged nearest neighbor classifier, so whenever NN does well, SVM should do even better, by adding automatic quality control over the examples in your set. On the converse -- if it's a dataset where nearest neighbor (in feature space) is expected to do badly, SVM will do badly as well.
- Is there a better unsupervised version of SVMs?
Just answering only this question here. Unsupervised learning can be done by so-called one-class support vector machines. Again, similar to normal SVMs, there is an element that promotes sparsity. In normal SVMs only a few points are considered important, the support vectors. In one-class SVMs again only a few points can be used to either:
"separate" a dataset as far from the origin as possible, or
define a radius as small as possible.
The advantages of normal SVMs carry over to this case. Compared to density estimation only a few points need to be considered. The disadvantages carry over as well.
Are SVMs better than ANNs with many classes?
SVMs have been designated for discrete classification. Before moving to ANNs, try ensemble methods like Random Forest , Gradient Boosting, Gaussian Probability Classification etc
What about in a semi-supervised case like reinforcement learning?
Deep Q learning provides better alternatives.
Is there a better unsupervised version of SVMs?
SVM is not suited for unsupervised learning. You have other alternatives for unsupervised learning : K-Means, Hierarchical clustering, TSNE clustering etc
From ANN perspective, you can try Autoencoder, General adversarial network
Few more useful links:
towardsdatascience
wikipedia