The AI world is changing faster than you think, so you have to keep your finger on the pulse. We have 5 links for you that are worth reviewing and information about events worth going to. Let's start reading the news about ML and applied AI!
5 Links
1.
Google has released the source code of the MorphNet neural network. A new library of neural networks to classify images that does faster. Read more on the
Google blog.
2. One of the most interesting applications of Machine Learning
techniques are clustering techniques. Vincent Granville has prepared a
list of 34 articles and tutorials on clustering.
3. During creating Machine Learning models, you need to analyze parameters for algorithms. Jakub Czakon wrote a very interesting
series of articles on
how to optimize hyperparameters in Python.
4. We have new solutions for
supporting .NET applications using the advantages of Spark cooperation with the opportunity to use the Azure ML solution. An interesting way to present a new panel for workflow ML projects in Spark -
MLFlow.
5.
The Pandas library is a standard in data analysis. The Pandas library is well-written for acting on small projects. Currently, Spark has been running a project Koalas, which based on libraries Pandas has implemented a program for data analysis in distributed environments works on multiple clusters Spark. You can watch the premiere of this library
Koalas Open Source Project from his creator Reynold Xin.