Headstart 2019

Welcome to the University of York 2019 Headstart programme website for the computer science session "I can see what you see - Looking through the eyes of a computer". On this page you will find links to the resources presented on the day and some interesting links which you may wish to investigate after the session ends.

If you want to continue investigating machine learning them I would recommend two books in particular:

Hands-On Machine Learning with Scikit-Learn and TensorFlow is a very good introduction to machine learning and, as the title suggests, has many practical exercises for the reader to follow.

Deep Learning with Python is more focused on neural networks but is a very clear introduction to convolutional neural networks which play a key role in image classification tasks.

A github page has been created from which all of the materials are available for download.

if you don't have a github account you can download most of the materials directly from the table below.

Link Description
Worksheet A PDF of the student worksheet including exercises for image classification, function fitting, convolution kernels and neural networks.
Berkeley Vision Website A demonstration of image classification using Caffe written at Berkeley.
Tensorflow Playground Website for the Tensorflow plaground.
Excel images Matt Parker's brilliant maths website with a tool to convert images to spreadsheets.
Excel and Convolutions Towards Data Science - Understanding convolutions using Excel
Processing Processing is a flexible software sketchbook. I think Processing is fantastic for teaching students how to code. It is based on java and allows rapid construction of creative programs. It was used to develop the demonstration software for this workshop.
Function Fitting The code for the processing application used to fit functions to data.
Exploring Kernels The code for the processing application used to explore the effects of convolution kernels on images.