Instant translations 📚, Azure-powered healthcare 👨‍⚕️, and the AI Revolution in Singapore 🤖

3 min read

What’s happening in Data? 

Here are the updates from AWS Machine Learning, Microsoft Azure, and Google Cloud.  

AWS Machine Learning 

AWS Deep Learning Challenge sees innovative and impactful use of Amazon EC2 DL1 instances: participants from academia, startups, and enterprise organizations joined to test their skills and train a deep learning model of their choice using Amazon Elastic Compute Cloud (Amazon EC2) DL1 instances and Habana’s SynapseAI SDK.

Build a multi-lingual document translation workflow with domain-specific and language-specific customization: In the digital world, providing information in a local language isn’t novel, but it can be a tedious and expensive task. Advancements in machine learning (ML) and natural language processing (NLP) have made this task much easier and less expensive.

Microsoft Azure 

Microsoft launches Azure Health Data Services to unify health data and power AI in the cloud: take a look at how the medical PaaS service is empowered by AI/ML and how you could help in the emerging field.

Hierarchical forecasting for Azure Machine Learning is now generally available: take a full look at the hierarchal forecasting available with Azure, right on GitHub.

Google Cloud 

Google Cloud and Singapore government team up on AI: Singapore’s National AI office will tap Google Cloud’s expertise in artificial intelligence to build artificial intelligence applications and train public sector officers on AI.


Understanding Machine Learning Algorithms

By Imran Ahmad

Understanding supervised machine learning

Formulating supervised machine learning

Chapter 7

Before going deeper into the details of supervised machine learning algorithms, let’s define some of the basic supervised machine learning terminologies:

A trained supervised machine learning model is capable of making predictions by estimating the target variable based on the features.

Let’s introduce the notation that we will be using in this chapter to discuss the machine learning techniques:

Now, let’s see how some of these terminologies are formulated practically.
As we discussed, a feature vector is defined as a data structure that has all the features stored in it.
If the number of features is and the number of training examples is b, then
X_train represents the training feature vector. Each example is a row in the feature vector.

Now, let’s assume that there are training examples and testing examples. A particular training example is represented by (Xy).
We use superscript to indicate which training example is which within the training set. So, our labeled dataset is represented by D = {X(1),y(1)), (X(2),y(2)), ….. , (X(d),y(d))}.We divide that into two parts—Dtrain and Dtest.

So, our training set can be represented by Dtrain = {X(1)  , y(1)), (X(2), y(2)), ….. , (X(b), y(b))}.

The objective of training a model is that for any ith example in the training set, the predicted value of the target value should be as close to the actual value in the examples as possible. In other words:

So, our testing set can be represented by Dtest = {X(1)  , y(1)), (X(2), y(2)), ….. , (X(c), y(c))}.

The values of the target variable are represented by a vector, Y: Y ={ y(1), y(2), ….., y(m)}

Read Here


Secret Knowledge: Building your Data Arsenal 

  • /numpy – The fundamental package for scientific computing with Python.
  • cupy/cupy – A NumPy/SciPy-compatible array library for GPU-accelerated computing with Python.
  • /numba – NumPy aware dynamic Python compiler using LLVM.