technology
In the Internet of Things era, a large number of perceptrons are collecting and generating data involving various fields every day. Due to the demand of improving the quality of business and life, it is a very valuable research direction to analyze big data streams by using Internet of Things technology. This paper outlines in detail the use of deep learning to improve data analysis and learning methods in the field of Internet of Things. From the perspective of machine learning, the author divides the methods of processing Internet of Things data into big data analysis and streaming data analysis. This paper summarizes the different deep learning methods at present, and discusses in detail the advantages of using deep learning methods to analyze Internet of Things data, as well as the challenges in the future.
In this series of articles, we introduced deep learning and long-term and short-term memory (LSTM) networks, showed how to generate data for anomaly detection, and also introduced how to use the Deeplearning4j toolkit. In this paper, we will introduce how ApacheSystemML, an open source machine learning system, helps to perform linear algebraic operations by dynamically optimizing execution and using ApacheSpark as a runtime engine. It also shows that the performance of even a very simple single-layer LSTM network is better than the advanced anomaly detection algorithm on time series sensor data (or any type of general sequence data).
GoogleAssistant and other natural language understanding platforms are pushing users to use their technology. Whether it is a simple task such as setting a timer or a more complicated task such as adjusting the thermostat with Google Assistant, you can participate. In this article, step by step, how to build your own assistant application is introduced, and plants are watered by simply asking Google to control AndroidThings devices.
Open source
Tinyweb is a simple lightweight HTTP server, which is used in micro devices such as ESP8266/ESP32 running MicroPython. Having a simple HTTP server allows developers to create beautiful and modern user interfaces for their IoT devices. Tinyweb itself is a simple TCP server running on uasyncio.
Mynewt is a component open source operating system suitable for micro embedded devices. ApacheMynewt uses newt to build a package management system, which allows developers to select only the components needed to build an operating system. Its goal is to make the application and development of microcontroller environment driven by power consumption and cost easy. Mynewt provides open source Bluetooth 5.0 protocol stack, embedded middleware, flash file system, network stack, bootloader, FATFS, bootloader, statistics and recording infrastructure.
AngularIotDashboard is a dashboard in the field of Internet of Things based on Angular4. It is a real-time compatible dashboard suitable for any browser, and its goal is to become an elastic front end of smart home, smart office and industrial automation. Through many reusable components, developers can inspire and implement their own version of the managed IoT dashboard based on AngularIoTDashboard.
Hardware appliance
FemtoUSB is an open source ARM development board based on ATSAMD2 1E 18A of ARMCortexM0+ products of Atmel. It aims to provide a basic starting point for those who are interested in ARM design, especially those who are ready to switch from AVR8-bit hardware to very powerful ARM32-bit tools. Its circuit board design, schematic diagram and parts list are completely open source, which allows developers to learn to design ARM chips, compile tool chains, basic circuit diagrams of ARM chips and so on.