Hands-On Machine Learning for Algorithmic Trading

Sensors

Data generated by networked sensors embedded in a broad range of devices are among the most rapidly growing data sources, driven by the proliferation of smartphones and the reduction in the cost of satellite technologies.

This category of alternative data is typically very unstructured and often significantly larger in volume than data generated by individuals or business processes, and poses much higher processing challenges. Key alternative data sources in this category include:

• Satellite imaging to monitor economic activity, such as construction, shipping, or commodity supply
• Geolocation data to track traffic in retail stores, such as using volunteered smartphone data, or on transport routes, such as on ships or trucks
• Cameras ...