Posted on: Monday, November 23rd, 2015
Let’s face it: Big Data is a big deal. In fact, it’s such a big deal that it was the main topic at the last Charlotte IoT meeting; during this meeting, one of the platforms discussed was Microsoft Azure. Azure is becoming known as a top tool for streaming analytics, machine learning and predictive analytics.
So what’s the deal with Microsoft Azure Machine Learning? Here are 10 things you need to know:
1) It’s a cloud computing service.
2) Almost all pricing is À la carte, meaning that, when it comes to features, you only pay for what you use.
3) The more data an algorithm has, the more accurate the predictions are.
4) Azure makes it easy to create experiments to help select the best machine learning algorithm.
5) Machine learning models are only as good as the data. Cleaning and filtering the data may hurt the algorithm’s ability to predict.
6) Azure Event Hubs can be used to ingest data from millions of concurrent connected devices.
7) Experiments for data models can be published in minutes (It used to take expert data scientists days).
8) Microsoft is using machine learning with the Clutter feature in Office 365. The more you use it, the more data the model has to make its predictions. Dan Thyer has been training his model* for months, and it’s uncanny how good the algorithm is at filtering out the lower priority emails to read. Out of 400 clutter emails last week, the algorithm only had one incorrectly filtered email.
9) Hadoop can be run on Azure. Hadoop is an open-source software framework for storing data and running applications on clusters of commodity hardware. It provides massive storage for any kind of data, enormous processing power and the ability to handle virtually limitless concurrent tasks or jobs.
10) Triggers can be built for Logic Apps to easily send notifications or texts without a bunch of code.
Microsoft Azure is making huge steps to making stream analytics, machine learning and predictive analytics more accessible. Before now, only the most extreme expert data scientists were able to work in this space, but now the technology is becoming more accessible to the masses!
*One example of how machine learning tools can be used in everyday life will soon be displayed in Dan Thyer’s home. Thyer, Logical Advantage’s CTO and Co-Founder, will build an IoT device with IoT to send data to the cloud where Azure Machine Learning will be used to predict if his pets are sick based on their eating habits. First, using a gate to their food and LFRFID tags on the animals, he’ll ensure that each animal only eats the food intended for it, meaning that the dog will no longer be able to eat the cat’s food! With the data he receives from this machine, he’ll be able to build a model around their normal eating habits. Some machine learning algorithms are good at seeing when data is not normal and making predictions on that data; for Thyer’s invention, this algorithm will send him a notification when it thinks something is wrong with an animal. This concept is similar to how machine learning is being used to predict when failures are about to occur in oil pumps and other equipment.
Stay tuned for more news on Microsoft Azure and Big Data platforms, and be sure to tweet your thoughts and questions to us @LogicalAdv!Go Back