confira os treinamentos programados para a próxima semana.
consulte sempre um engenheiro eletrônico
In this course we will look at the types of analysis appropriate for the Internet of Things (IoT) and the analyses that are appropriate for each type. As a follow on to the course “The Internet of Things (IoT) and Analytics“, we will concentrate on the analytics and algorithms involved. We will also discuss an overview of the platforms available to implement these algorithms. Some examples of the types of analyses we will cover include failure prediction, optimization, and anomaly detection. We will give a flavor of the statistical methods used and their implementation.
- September 12 – Day 1: Introduction to Analytics for the IoT
Today we will present an overview of the types of statistical techniques useful for the IoT and how statistics is used in this context. Some differences with standard statistics will be highlighted. The requirements for IoT analytics will also be presented. Our intent is to discuss the algorithms in the context of engineering applications in the IoT.
- September 13 – Day 2: Languages and Environments
To implement the analytics we are discussing this week, one needs software and a platform to run it on. We will discuss languages such as R and Python and more conventional languages and traditional statistical processing systems. Because the size of the data is often massive and growing, platforms become important. It is key to pick one that can start small and grow in power without having to completely rewrite the application. We will give an overview of platforms specific to the IoT.
- September 14 – Day 3: Classification
A major area of machine learning concerns classification. We will describe both supervised and unsupervised learning techniques and discuss when and how to use them. Techniques include k-means clustering, support vector machines, principal components analysis and decision trees, for example. We will also look at applications of these techniques.
- September 15 – Day 4: Failure Analysis & Anomaly Detection
For this lecture we will talk about failure detection and anomaly detection, and the analyses used in conjunction with them. In the IoT we often want to perform these on a continuous basis. This introduces some interesting features in the analysis that we will discuss.
- September 16 – Day 5: Optimization
One major feature of the IoT is that we now have lots of data over long periods of time for many types of systems.We can use this rich data horde to perform optimizations in current systems and in the design of new or evolved systems.These techniques often use complex formulations involving linear and non-linear systems approaches.Fortunately, there are many existing implementations of these methods that make their execution relatively easy.We will look at them in a couple different languages.
- Simulating systems, evaluating LEDs, and specifying optics
- How to Never Re-do Your Documentation Again
- Understanding Flame Detection: How to Protect Your Facility from Fire Hazards
Clique na figura abaixo para ver a agenda completa com os links para as inscrições.