Information Center

Five ways that machine learning affects modern cloud computing

  

The cloud computing industry is gradually changing to the direction of intelligence. Although computing, storage and networking are still the main revenue sources of cloud providers, machine learning is slowly becoming the focus of contemporary cloud computing.

The following are five cloud services affected by machines:

Machine learning _meitu_2

Cognitive computing

The goal of cognitive computing is to enable applications to have the ability to see, listen, speak, and even make decisions. Based on natural language processing, visual recognition, facial recognition, emotion recognition, video analysis, text to speech, voice transcoding, language translation, emotion analysis and other technologies, cognitive computing enables developers to develop programming through simple APIs. By leveraging these services, applications can provide a more natural user experience. All these are behind machine learning, and various algorithms are applied to provide these powerful cognitive abilities. Although it seems simple, cloud providers have invested a lot of resources to provide cognitive APIs for developers. From insurance to financing, vertical industries in all major industries will start to use cognitive computing platform to provide better experience for customers.

Amazon AI, IBM Watson, Google Cloud and Microsoft Cognitive API are common commercial products in the market.

Robot service (robot service)

As the use of mobile applications stagnates, various businesses have begun to turn to interactive robots to promote customer service and support. Robots are rapidly replacing applications by providing a conversation experience for customers. For platforms such as microphones, whatsApp, Facebook Messenger and Slack, the demand for implantable robots is increasing. The introduction of robot concept on the platform began with Yahoo chat, but the application of machine learning makes it more valuable. Now, developers can use the old conversation mode to train robots. In addition to answering standard questions, robots can also have meaningful discussions with users.

In this regard, some emerging platforms are API.ai, IBM Watson Botkit and Microsoft Azure Bots modification, etc.

Internet of Things

Although the Internet of Things has taken different forms in the past 20 years, the data driven cloud platform still redefines this trend. In addition to capturing a large number of sensors from various data for query, it can also process and analyze various important trends. Machine learning can make cloud computing more intelligent. Predictive maintenance is a more attractive use case. In industrial networks, such a platform can replace the equipment for human fault monitoring. Various machine learning algorithms can work in series and evolve into a suitable mode to best understand the data set mode generated by the device. These operation modes can actively detect abnormalities that may eventually lead to equipment downtime, which will enable the industrial network to enter the next stage.

Two typical examples of predictive maintenance solutions for the Internet are Microsoft Azure IoT Suite and IBM Watson IoT.

Personal assistant

The rise of machine learning makes personal voice assistant more important than before. These assistants can learn about your past choices and use trends to provide you with customized application experiences. For example, it only takes a few days to create a playlist that best adapts to your emotional tendencies dynamically. They become smarter when notified and reminded. The APIs presented by these assistants will allow developers to master the power of ML (machine language). They can bring users a deeper customization experience.

Some intelligent personal assistants that provide technical support through machine learning are Amazon Alexa, Apple Siri, Google Assistant and Microsoft Cortana.

business intelligence

Traditional data warehouses have been plagued by a large amount of data and Apache Hadoop, and machine learning has been introduced into enterprise data warehouses. Decision makers can gain more intelligent insights from existing data and more accurately predict business trends. Including SCM, CRM, ERP, MRP, HR, sales and finance, will benefit from ML driven observation.

Yayamason, Google, IBM and Microsoft are building a bridge between traditional business intelligence platforms and emerging ML tools. Amazon Kinesis Analytics owns Amazon ML, Azure Stream Analytics owns Azure ML web services, and Google also uses Cloud ML to make BigQuery and Cloud DataFlow easier. Developers and architects can easily connect to the network and develop next-generation business intelligence tools.

These examples illustrate how machine learning has become the focus of intelligent cloud computing. In the next few quarters, we will see additional services and use cases provided by cloud providers.