video
The 5 best programming languages for AI development
Here are the top five languages for creating AI-enabled apps, according to InfoWorld contributor and deep learning architect Ian Pointer.
Similar
Comments
Who cares most about AI? Developers!
They’re the ones creating software that uses machine learning to detect patterns and help anticipate user actions, improve fraud detection, or enhance image or speech recognition.
Here are the top five languages for creating AI-enabled apps, according to InfoWorld contributor and deep learning architect Ian Pointer.
No. 5: The R language
Data scientists love R. But others may find it confusing – and the performance and operational hurdles are too high for production
• If you have a dedicated group of R developers, they can use R integrations with TensorFlow, Keras, or H2O…but…
• R should be for research, prototyping, and experimentation only—recode your R prototype in Java or Python!
No. 4: JavaScript
JavaScript is poised to democratize the use of AI, thanks to:
• TensorFlow.js, a WebGL-accelerated library for training and runing machine learning models in your web browser, which includes…
• The Keras API and the ability to use models that were trained in regular TensorFlow
Downside: JavaScript lacks the access to machine learning libraries enjoyed by other languages
No. 3: C /C++
If you’re writing code for an embedded app and need top performance in a small footprint, C/C++ is the answer. You can:
• Use libraries like CUDA to write your own code that runs directly on your GPU
• Use TensorFlow or Caffe to access flexible high-level APIs, letting you import models your data scientists may have built with Python
No. 2: Java and friends
The JVM family of languages
Java, Scala, Kotlin, Clojure
is also a great choice for AI application development. Libraries include
• CoreNLP – Natural language processing
• ND4J – Numerical operations including tensor
• DL4J – GPU-accelerated deep learning stack
Also, the Java family plays well with Apache Spark, the most popular framework for big data processing and machine learning.
No. 1: Python
Python is at the forefront of AI research and more AI libraries abound for Python than for any other language.
NumPy – Almost a standard API for tensor operations
Pandas – Delivers R’s powerful and flexible dataframes
NLTK and SpaCy – Natural language processing
Scikit-learn – Machine learning with data mining and analysis
Plus, all the popular deep learning libraries are effectively Python-first projects.
TensorFlow, PyTorch, Chainer, Apache MXNet, Theano
They’re the ones creating software that uses machine learning to detect patterns and help anticipate user actions, improve fraud detection, or enhance image or speech recognition.
Here are the top five languages for creating AI-enabled apps, according to InfoWorld contributor and deep learning architect Ian Pointer.
No. 5: The R language
Data scientists love R. But others may find it confusing – and the performance and operational hurdles are too high for production
• If you have a dedicated group of R developers, they can use R integrations with TensorFlow, Keras, or H2O…but…
• R should be for research, prototyping, and experimentation only—recode your R prototype in Java or Python!
No. 4: JavaScript
JavaScript is poised to democratize the use of AI, thanks to:
• TensorFlow.js, a WebGL-accelerated library for training and runing machine learning models in your web browser, which includes…
• The Keras API and the ability to use models that were trained in regular TensorFlow
Downside: JavaScript lacks the access to machine learning libraries enjoyed by other languages
No. 3: C /C++
If you’re writing code for an embedded app and need top performance in a small footprint, C/C++ is the answer. You can:
• Use libraries like CUDA to write your own code that runs directly on your GPU
• Use TensorFlow or Caffe to access flexible high-level APIs, letting you import models your data scientists may have built with Python
No. 2: Java and friends
The JVM family of languages
Java, Scala, Kotlin, Clojure
is also a great choice for AI application development. Libraries include
• CoreNLP – Natural language processing
• ND4J – Numerical operations including tensor
• DL4J – GPU-accelerated deep learning stack
Also, the Java family plays well with Apache Spark, the most popular framework for big data processing and machine learning.
No. 1: Python
Python is at the forefront of AI research and more AI libraries abound for Python than for any other language.
NumPy – Almost a standard API for tensor operations
Pandas – Delivers R’s powerful and flexible dataframes
NLTK and SpaCy – Natural language processing
Scikit-learn – Machine learning with data mining and analysis
Plus, all the popular deep learning libraries are effectively Python-first projects.
TensorFlow, PyTorch, Chainer, Apache MXNet, Theano
Featured
