Experiment and Result Analysis

Introduction

Here I describe the methodologies used to analyse the incident detection learning methods from the machine learning perpective.

Result Analysis

I have described the Experiment and the results in the last chapters (5.2 - 6) of my dissertation. A quick overview of the Area Under Curve values in Hong Kong roads with different datasets can be seen on this post.

Methods to be used - TODO

• SSE
• Learning Curves -
• Pipeline analysis - what part to focus on.
• Analysing High variance vs Bias
• Accuracy, Precision, Recall

References

https://www.coursera.org/learn/machine-learning/home/welcome