Machine Learning using SQL, a Visual Python Debugger, and more Machine Learning Developments!
A lot of aspiring data scientists these days are coming from a software engineering background so they are familiar with SQL. Thanks to Google, that’s all you might need to build machine learning models! With BigQuery ML, the tech giant is looking to help SQL users get integrated into the ML domain. More details in this article.
Other highlights from the past week: Researchers from IIT Hyderabad are using machine learning to identify bikers driving without helmets, IBM have pioneered an algorithm that adds a watermark to your neural network models, and more updates below.
Click on the headline to read the full article.
- Making AI Systems Faster and More Efficient with Google Edge TPUs: Ever since it’s introduction in 2016, TPUs have seen their demand rise as more complex data-driven problems have come up. And now the tech giant has launched Edge TPU, a small artificial intelligence accelerator that enables machine learning jobs in IoT (Internet of Things) devices!
- Google’s BigQuery ML Tool Enables you to do Machine Learning through SQL!: Are you a SQL user looking to get into data science? Great news! Google has launched a platform called BigQuery ML that lets you do machine learning using SQL queries.
- PixieDebugger — A Visual Python Debugger for Jupyter Notebooks Every Data Scientist should Use: If you use Jupyter Notebooks, PixieDebugger is a MUST USE tool. It’s a visual debugger for your python code and it is packed full of awesome features. Once you start using it, you won’t know how you worked without it!
- IIT Hyderabad Researchers are using Machine Learning to Identify Bike Drivers without Helmets: Researchers from IIT Hyderabad have designed a computer vision algorithm (using convolutional neural networks) that detects bike riders without helmets and notifies the police. This technology could soon be in action on the streets!
- Protect your Machine Learning Model with IBM’s Watermark Algorithm: Data scientists spend so much time building models, shouldn’t they be able to protect them from misuse? IBM’s latest research helps data scientists do just that by adding watermarks to their models! This article looks at how IBM’s algorithm works.
The above AVBytes were published from 23rd to 29th July, 2018.