New open-source Machine Studying Framework written in Java

I’m joyful to announce that the Datumbox Machine Studying Framework is now open sourced beneath GPL 3.0 and you may obtain its code from Github!

What is that this Framework?

The Datumbox Machine Studying Framework is an open-source framework written in Java which allows the fast improvement of Machine Studying fashions and Statistical purposes. It’s the code that presently powers up the Datumbox API. The principle focus of the framework is to incorporate a lot of machine studying algorithms & statistical strategies and have the ability to deal with small-medium sized datasets. Although the framework targets to help the event of fashions from varied fields, it additionally supplies instruments which are notably helpful in Pure Language Processing and Textual content Evaluation purposes.

What sorts of fashions/algorithms are supported?

The framework is split in a number of Layers comparable to Machine Studying, Statistics, Arithmetic, Algorithms and Utilities. Every of them supplies a sequence of lessons which are used for coaching machine studying fashions. The 2 most necessary layers are the Statistics and the Machine Studying layer.

The Statistics layer supplies lessons for calculating descriptive statistics, performing varied sorts of sampling, estimating CDFs and PDFs from generally used likelihood distributions and performing over 35 parametric and non-parametric exams. Such sorts of lessons are often mandatory whereas performing explanatory knowledge evaluation, sampling and have choice.

The Machine Studying layer supplies lessons can be utilized in a lot of issues together with Classification, Regression, Cluster Evaluation, Subject Modeling, Dimensionality Discount, Characteristic Choice, Ensemble Studying and Recommender Programs. Listed below are a number of the supported algorithms: LDA, Max Entropy, Naive Bayes, SVM, Bootstrap Aggregating, Adaboost, Kmeans, Hierarchical Clustering, Dirichlet Course of Combination Fashions, Softmax Regression, Ordinal Regression, Linear Regression, Stepwise Regression, PCA and extra.

Datumbox Framework VS Mahout VS Scikit-Be taught

Each Mahout and Scikit-Be taught are nice tasks and each of them have fully completely different targets. Mahout helps solely a really restricted variety of algorithms which could be parallelized and thus use Hadoop’s Map-Cut back framework to deal with Large Information. Then again Scikit-Be taught helps a lot of algorithms however it may well’t deal with large quantity of information. Furthermore it’s developed in Python, which is a superb language for prototyping and Scientific Computing however not my private favorite for software program improvement.

The Datumbox Framework sits in the midst of the 2 options. It tries to help a lot of algorithms and it’s written in Java. Because of this it may be integrated simpler into manufacturing code, it may well simpler be tweaked to cut back reminiscence consumption and it may be utilized in actual time programs. Lastly despite the fact that presently Datumbox Framework is able to dealing with medium-sized datasets, it’s inside my plans to broaden it to deal with large-sized datasets.

How steady is it?

The early variations of the framework (as much as 0.3.x) have been developed in August and September of 2013 they usually have been written in PHP (yeap!). Throughout Might and June 2014 (variations 0.4.x), the framework was rewritten in Java and enhanced with further options. Each branches have been closely examined in industrial purposes together with the Datumbox API. The present model is 0.5.0 and it appears mature sufficient to be launched as the primary public alpha model of the framework. Having stated that, it is very important be aware that some functionalities of the framework are examined extra completely than others. Furthermore since this model is alpha, you need to count on drastic adjustments on the longer term releases.

Why I wrote it and why I open-source it?

My involvement with Machine Studying and NLP dates again to 2009 once I co-founded WebSEOAnalytics.com. Since then I’ve been creating implementations of varied machine studying algorithms for varied tasks and purposes. Sadly a lot of the authentic implementations have been very problem-specific they usually might hardly be utilized in another downside. In August 2013 I made a decision to start out Datumbox as a private undertaking and develop a framework that gives the instruments for creating machine studying fashions focusing within the space of NLP and Textual content Classification. My goal was to construct a framework that will be reused on the longer term for creating shortly machine studying fashions, incorporating it in tasks that require machine studying parts or provide it as a service (Machine Studying as a Service).

And right here I’m now, a number of strains of code later, open-sourcing the undertaking. Why? The sincere reply is that at this level, it isn’t inside my plans to undergo a “let’s construct a brand new start-up” journey. On the similar time I felt that maintaining the code on my arduous disk in case I want it on the longer term doesn’t make sense. So the one logical factor to do was to open-source it. 🙂

Documentation?

If you happen to learn the earlier two paragraphs, you need to most likely seen this coming. For the reason that framework was not developed having in thoughts that I’d share it with others, the documentation is poor/non-existent. Many of the lessons and public strategies will not be correctly commented and there’s no doc describing the structure of the code. Luckily all the category names are self-explanatory and the framework supplies JUnit exams for each public methodology & algorithm and these can be utilized as examples of the way to use the code. I hope that with the assistance of the group we are going to construct a correct documentation, so I’m relying on you!

Present Limitations and Future Improvement

As in each piece of software program (and particularly the open-source tasks in alpha model), the Datumbox Machine Studying Framework comes with its personal distinctive and lovable limitations. Let’s dig into them:

1. Documentation: As talked about earlier, the documentation is poor.
2. No Multithreading: Sadly the framework doesn’t presently help Multithreading. After all we should always be aware that not all machine studying algorithms could be parallelized.
3. Code Examples: For the reason that framework has simply been revealed, you may’t discover any code examples on the net apart from these offered by the framework within the type of JUnit exams.
4. Code Construction: Making a strong structure for any giant undertaking is at all times difficult, not to mention when you need to take care of Machine Studying algorithms that differ considerably (supervised studying, unsupervised studying, dimensionality discount algorithms and many others).
5. Mannequin Persistence and Giant Information Collections: Presently the fashions could be skilled and saved both on recordsdata on disk or in MongoDB databases. To have the ability to deal with great amount of information, different options have to be investigated. For instance MapDB looks like candidate for storing knowledge and parameters whereas coaching. Furthermore it is very important take away any 3^rd occasion libraries that presently deal with the persistence of the fashions and develop a greater dry and modular resolution.
6. New algorithms/exams/fashions: There are such a lot of nice strategies that aren’t presently supported (particularly for time sequence evaluation).

Sadly all of the above are an excessive amount of work and there may be so little time. That’s the reason in case you are within the undertaking, step ahead and provides me a hand with any of the above. Furthermore I’d love to listen to from individuals who have expertise in open-sourcing medium-large tasks and will present any recommendations on the way to handle them. Moreover I’d be grateful to any courageous soul who would dare to look into the code and doc some lessons or public strategies. Final however not least should you use the framework for something attention-grabbing, please drop me a line or share it with a weblog publish.

Lastly I want to thank my love Kyriaki for tolerating me whereas scripting this undertaking, my buddy and super-ninja-Java-developer Eleftherios Bampaletakis for serving to out with necessary Java points and also you for getting concerned within the undertaking. I’m wanting ahead to your feedback.