scala

Adam is one more optimization algorithm used in neural networks. It is based on adaptive estimates of lower-order moments. It has more hyper-parameters than classic Gradient Descent to tune externally

Good default settings for the tested machine learning problems are:

• α = 0.001, // learning rate. We have already seen this one in classic Gradient Descent.
• β₁ = 0.9,
• β₂ = 0.999
• eps = 10−8.

In this article we are going to use Scala mini-library for Deep Learning that we developed earlier in order to study basic linear regression task. We will learn model weights using perceptron model, which will be our single unit network layer that emits target value. This model will predict a target value yHat based on two trained parameters: weight and bias. Both are scalar numbers. Weights optimization is going to be based on implemented Gradient descent algorithm:

• Gradient Descent Optimization
• Training Loop

Model equation:

y = bias + weight * x

TensorFlow Scala is a strongly-typed Scala API for TensorFlow core C++ library developed by Anthony Platanios. This library integrates with native TensorFlow library via JNI, so no intermediate official/non-official Java libraries are used.

In this article we are going to implement ANN from scratch in Scala. It is continuation of the first article, which describes a theory of ANN.

This implementation will consist of:

If you need to run a Scala code as a script, i.e. using Scala source file to execute some short code, then Ammonite Scripts may be a solution for you. Ammonite project consists of a REPL, script launcher and a few Scala libraries.
Let’s write a script to generate JSON data and send it to Apache Kafka topic.

Most of the algorithmic tasks are related to iterating over arrays of data. They often can be expressed as a function which takes some input and returns some single value or an array of values. For instance:

def maxSum(a: Array[Int]): Array[Int] = ???

SBT is a Scala Build Tool. It is written in Scala and can compile, build artefacts for Scala and Java projects. SBT is also the first build tool in the Scala eco-system and the most used one among Scala developers. I am using SBT already for many years and found the following useful plugins which I use in most of my projects:

Once you start dig deeper into Scala and its suitability for functional programming, you meet Monads. In this blog post, we will explore Monads in Scala: their usage and usefulness.

Sometimes Scala developer needs to call external program, which is running outside of the JVM. In this case, we use scala.sys.process package. Process package has bunch of functions to spin up new processes, consume their outputs and errors. Also, spawned process can be stopped. Usually, we run external programs for a short period of time to make some side-effect. Then, we analyse its exit code to apply some error handling logic in our main Scala program. It worth to say that process API is blocking execution thread, when we are waiting for its completion. To summarise, Scala developer wants to do the following:

Recently, I ran into a familiar situation by doing data processing, where I needed to deal with a fragmented data stream. Having fragments, I had to detect manually where exactly new line/message starts and where current line/message ends in the stream. As turned out, one can aggregate intermediate state of the fragmented stream using scan function.

Let us dig down into how scan function is working.