Since 2016, Simon Fraser University’s Big Data Hub has offered workshops, events, and consulting services for researchers and industry

Simon Fraser University also hosts the Cedar supercomputer—a cluster of 100,400 CPUs and 1352 GPUs—soon to be replaced by an even bigger cluster

Goals for this workshop

Clarify important concepts around big data

Demonstrate how important big data has become in agriculture and farming

This workshop will be followed by a 2^{nd} session on challenges and solutions while working with big data

Concepts

Big data

Datasets so large that they cannot be processed by conventional analytics methods, but which open up new possibilities for predictive modelling and machine learning

Machine learning (ML)

A subfield of artificial intelligence that can be defined as computer programs whose performance at a task improves with experience

Modelling

A subfield of ML, modelling is the use of mathematical models to calculate predictions

Deep learning (DL)

A subfield of ML using artificial neural networks to calculate predictions or generate novel data

GPUs, initially designed to accelerate computer graphics, proved well suited to run vast amounts of simple computations in parallel

GPUs’ new usage in particular opened the door to deep neural networks

The power of deep neural networks

In typical computing, a programmer writes code that gives a computer detailed instructions of what to do

This works very well for a number of applications, but many tasks would require an impossible amount of code and resources

By feeding a very large amount of data to complex algorithms such as neural networks however, we can train them to become better and better at some tasks without explicitly programming how they should do so

Big data in agriculture

Big data road map

An integrated dynamic system

Wolfert, S., Ge, L., Verdouw, C., & Bogaardt, M. J. (2017). Big data in smart farming–a review. Agricultural systems, 153, 69-80.

Case studies

Resources

Understanding neural networks

3Blue1Brown by Grant Sanderson has a series of 4 videos on neural networks which is easy to watch, fun, and does an excellent job at introducing the functioning of a simple neural network