- US dollar rate
- Euro rate
- Yuan rate
- Oil price
Although there are some reasons I chose them, in these kinds of economic analysis especially in Machine Learning methods the more data you have, the better prediction you can have, and the above data were the only easy digestible I found. Before anything else let me say that:
The calculation shows it is more likely to get better than worse.
The CAD rate prediction for next 100 days
|
The learning factor or the amount of increasing the scores is based on the year of the data. The closer to the current year the more increase in the score. For example, in the year 2006, the learning factor is 6/12 and in 2015, it is 15/16 and in 2016, it is 16/16. The result was a graph like the following pictures:
Bayesian Network for the above processed information. |
The SVG format of the above network is available in the following too, so you can download it and have a high-quality view in your browser: "SVG format of Canadian dollar analysis network"
Then based on the given network of information, I wrote an application to traverse the network randomly based on existing probabilities. The application was not that easy to write, it required some forward, and backward movement on the network and this kind of action on a Bayesian Network is somehow tricky. I ran it 15 times and got 15 different behavior as the following:
CAD's 15 different scenarios for next 100 days |
As you see from this 15 scenarios, there are only three scenarios which give lower than 0.71, so the probability of going lower than 0.71 USD is about 20%.
While the likelihood of having the rate between 0.71 and 0.77 is about 53% and the best of all having the rate above 0.77 but not more than 0.83 is about 27%
And let me know if you have any question, send me a message from the right side column.
No comments:
Post a Comment