Machine Learning Updates from the Largest Tech Companies

Recent updates about machine learning innovation and developments from tech giants

Machine learning

Machine learning

Developments in machine learning algorithms have significantly grown over the past few years. The collection of statistical methods and algorithms can accelerate the growth and productivity of business processes and can also aid in drug discovery and applied research. We can successfully employ ML algorithms for classification, regression, and assembling high-dimensional data.

Below are some of the recent updates by the top tech giant companies about their recent developments, innovations, and reports on machine learning.


Apple continues to refine Siri’s location-based responses:

Recently, Apple announced that the company will continue to work to bring precision to Siri’s location-based responses. The development will allow more refinement and location-aware voice commands. Apple aims to achieve the goal that a single digital assistant will perform on various devices; instead of multi virtual assistants. The company has received a patent for ‘digital assistant hardware abstraction’ that will allow the team to connect context sharing and task performances; across groups of devices with intelligent assistant capabilities.


Databricks launches open-source data sharing protocol:

Databricks has launched an open-source data sharing protocol to ensure maximum security while transferring data across different organizational platforms in real-time. Delta Sharing, the latest open-source protocol, enables a common standard for sharing structured and unstructured data, with a protocol that can be used in SQL, visual analytics tools, and various programming languages like Python and R. Delta sharing extends the capability of the delta lakehouse architecture, making it more simple, open and collaborative with machine learning.


IBM launches AI toolkit for deep-learning unpredictability

IBM released an open-source toolkit that supports AI uncertainty. Uncertainty Quantification 360 (UQ360) aims to provide data scientists and developers machine learning algorithms to improve ML models through quantification, evaluation, and effective communication. This toolkit is designed to enable human interference in automated systems. It analyzes different types of AI uncertainties using ML algorithms and helps to work on those issues.


Amazon starts ML Summer School

Amazon announces the launch of an integrated learning program for students who are interested to learn applied machine learning skills and are looking to develop their careers in the technology sector. The program has been initiated to train students and meet the growing demands for such skills across several industries. Candidates will be verified through online assessments. They will earn the first-hand experience in ML techniques like deep-learning and probabilistic graphical models to solve complex business problems.


Google’s AI-designed chip tells us about the nature of intelligence:

Even though the use of software in chip designing is old, Google’s researchers claim that the new reinforcement model automates the generation of chip floor plans which are superior or equivalent to those produced by humans in key metrics like power consumption, performance, and chip area.


ISRO is offering online courses on machine learning and GIS technology

ISRO is currently offering short courses in ML, GIS technology, and Earth Observation for Carbon Cycle Studies. The organization is extending a chance to directly interact with the reputed scientists working there and learn from them. This course is specifically designed for professionals working in remote sensing data processing in different applications.


dotData introduces a new toolkit to make machine learning useful

Implementing the right features in an AI algorithm is essential for businesses requires a greater understanding for business leaders to ensure efficient decision making. dotData aims to automate this entire process. The company’s new dotData Py Lite is a packed-up version of their ML toolkit which allows the users to efficiently build proofs of concept (POCs). Users can either download the toolkit or run it through dotData’s cloud service.