The first version of the system will be organized as follows:
Variables
Load data
Preprocess
Forecast each data stream
Optional: sensitivity analysis
Create a portfolio
Calculate position size
Repeat (weekly?)
Here are the stub Matlab files. This is basically a shell to test each piece of the system as they are created. Sys1.m is the script that runs each step, however steps can be run individually. Each step is implemented as a function with certain inputs and outputs rather than as a series of scripts. I am open to suggestions for refactoring.
The loaddata.m file is the longest so far- it pulls data from Yahoo finance. For forecasting I will probably use libSVM.
Lastly, here is a list of functions that looked like they will be useful. Most are from the statistics and financial toolboxes; svms are in the bioinformatics toolbox but I’ll probably use libSVM instead since it’s familiar to me.
![Introducing the Knowledge Worker [VIDEO]](https://www.smartdatacollective.com/wp-content/uploads/2026/08/introducing-knowledge-worker-video-1-330x220.jpg)
Please feel free to suggest or criticize anything.
