How to solve example 5.3 step per step?

[qwertyqwertz]

Hi folks,

I wonder how to solve the example 5.3 on page 174? I tried to come up with the right solution, but the outcome doesn't make sense.

These are the given numbers:
Hitrate: 52.1%
Number of test days: 500

The correct result is a cumulative profit of 22%.

How do I get 22%?

Thanks a lot and have a nice day!

[yaser]

Originally Posted by qwertyqwertz

Hi folks,

I wonder how to solve the example 5.3 on page 174? I tried to come up with the right solution, but the outcome doesn't make sense.

These are the given numbers:
Hitrate: 52.1%
Number of test days: 500

The correct result is a cumulative profit of 22%.

How do I get 22%?

Thanks a lot and have a nice day!

The hit rate is the percentage of days you got the direction right, while the profit takes into consideration how much money you made (or lost) when you got the direction right (or wrong). You can read the 22% directly off the figure on page 175 as the final value at day 500.

[qwertyqwertz]

Originally Posted by yaser

The hit rate is the percentage of days you got the direction right, while the profit takes into consideration how much money you made (or lost) when you got the direction right (or wrong). You can read the 22% directly off the figure on page 175 as the final value at day 500.

Hello Yaser,

Thank you very much for taking the time to answer my question! Somehow I didn't made the connection between that specific paragraph on page 174 and the figure on page 175 which is quite silly, because it follows right after.

In addition, let me thank you and your co-authors for your brilliant and concise book "Learning from Data". It was a delight to read! Though, the mathematical notation was sometimes a little odd for my taste. But this is just a minor critic that doesn't hurt the overall reading experience.

The book shipped without any problem from the Amazon US store to Germany. The delivery took about six weeks.

The only thing I found lacking from the book was a chapter about evaluation measures. I think this is a critical point that would help the reader understand why a certain model is better in some regards than another. Moreover, it would be really nice if you could provide some examples to illustrate the different measures since this is often quite dependent on the context in which the model in question is applied.

Again, let me thank for your excellent book and online course!

[yaser]

Thank you for the information and the feedback.