Machine Learning

The task list is getting chomped through

• Evaluate Dataset
• Choose Classification Method
  • Justify
• Implement Method
• Measure Accuracy
• Discuss Issues
• Discuss Reliability
• Suggest Improvements
  • Data gathering
  • Processing

I’m not a 100% sure how to justify my assumption that the result is linearly separable. Also I’m not sure a Perceptron is the best choice, but its easy to implement and within the scope of the module. The more complex methods I’m thinking of are outside the scope of the module, so I think its possible best to ignore them.

As for my validation method I’m thinking of using Leave-one-out cross-validation method to measure accuracy.

Still not really happy with my progress today because the render slowed down my laptop most of the morning. Which fractured my focus.

So how do I justify the assumption the data is linearly separable?