TensorFlow

What is TensorFlow?

TensorFlow is an open source library developed by Google which is widely used for machine learning applications.

TensorFlow is most commonly used with Python but is also available in a variety of programming languages.

Open-source software has been developed by a community of people and is made available in full for others to use, inspect, reuse and contribute to.

A library is a collection of files and functions allow the user to do tasks without having to code it themselves

Machine learning is the creation of applications that can automatically perform tasks, predict outcomes and develop insights in data through the use of probabilities, statistical models and algorithms. It is used in a variety of contexts, famous examples being Google’s personalised searching and Facebook’s news feed

What is TensorFlow useful for?

TensorFlow is used to develop machine learning applications. Instead of explicitly telling the computer what to do, machine learning uses statistical and mathematical models on large data sets to determine patterns, similarities and comparisons in the data. The application can then perform specific tasks based on these calculations. These tasks and calculations can be further developed and refined through the use of more data. This can be used to automatically perform tasks and provide predictions and insights into existing and new data.

Machine learning can be used to solve research problems in, and not limited to, image and text analysis. Take the example of images of animals. A machine learning application can be trained using the data of animal images with correct animal names to predict the names of animals in new images without names. It can also be used to analyse, develop insights into and explore the raw data of animal images without names by grouping and classifying them based on similarities and patterns identified by the machine learning application, possibly identifying new patterns unknown to the user.

Where can I get it?

Available free, visit TensorFlow for more information.

TensorFlow workshops

TensorFlow events run regularly throughout the year and include the following workshops:

  • Introduction to TensorFlow (Note: experience with Python is required for this workshop.)

Registering for a workshop

Workshop registration is available through myDevelopment. This aligns with myPlan and training records for staff, and enables more streamlined processes for assigning credit towards the Monash Doctoral Program for Graduate Research students.

To search for upcoming Data Fluency workshops log into myDevelopment from your myMonash portal. Type "Data Fluency" or the software you are interested in (eg. "python") into the search box to find a list of available workshops.

No access to myDevelopment?

If you do not have access to myDevelopment, please complete the relevant form:

Staff Development will validate and approve access. This process will take one business day to complete. An email will be sent to users confirming access details.

Where else to find training

You can find more online training materials for this tool via the Library. Visit LinkedIn Learning or Safari to access a range of videos, eBooks and online courses, or try using Library Search to find other resources to help you master this tool.

If you're still not sure where to start, use the details below to get in touch with the Data Fluency community.

Find out more

For further information or advice come to the weekly drop-in session, join the community discussion on Slack, subscribe to our mailing list or email datafluency@monash.edu.

.

TensorFlow

Can be used for:
  • Analysing data

View our workshop materials on GitHub