Mathias Bank

For a practical course I have used a Wacom tablet to write some characters. The goal is to recognize these handwritten characters with the help of a mlp.

So far, I have written the characters using my mouse. But this was not as fast and good as with this tablet. I have just created 29 Sample-Sets (each with 62 Samples) in just one hour. I think, that’s a fantastic job.

Perhaps I will buy such a tablet for my own work especially in graphics. This will surely improve my work .

Ok, I have achieved to create a small sample in Torch3. Cause of the bad documentation, I have just tried to build a multi layer perceptron for a xor-problem. As we should know, this problem can only be solved with at least one hidden layer and not by a simle perceptron.

Ok, the first problem was: how can I build a sample file. As I have seen in a tutorial, there exists a function to load simple ASCII-files. But there was no description on how to build them. But that’s really simple:

In the first line, there has to be to integers: the first specifies the number of examples, the second number specifies the number of datapoints in each example. Each sample exists of the input data and the output data. Each one is seperated by blank. The xor-problem can be seen in this sample-file. This file can be simple loaded:

MatDataSet *data = new(allocator) MatDataSet("XOR-ProblemData.txt", 2, 1);

The first parameter specifies the sample file. The second specifies the number of input data and the last number specifies the number of the output data. So in the xor-problem we have to inputs and one output.

And now, the complete source of our program. I think, it needn’t to be explained. If you have read the tutorial, it’s just easy to unterstand, what is done.

I hope, this short sample helps others who want to start with Torch3.

OpenSource is a really nice idea and helps to program things much faster. They help you not do invent things, which are already invented. But for me, they are mostly just a method to lern new programming paradigmas. What’s really annoying me is the documentation.

For a programming project for letter recognition at my university, I want to use Torch3, a library for machine learning in c++. The tutorial shows, that the library is really interesting. But unfortunatelly, the documentation is not very good. Only the examples help a little bit to use the library.

What can we learn? If we want to publish an open source project, the documentation is very very important. If we want that many people use our library, this is nearly as important as the library itself.

Ok, let’s try to use the library….

Some time ago, I have written a short summary about fuzzy logic. I think, now is a good time to make this paper public, so everyone can read it and it’s possible to see how we can use it.

Fuzzy logic is a nice possibilty if we don’t want or we can’t devide the world into white and black. It’s used in many disciplines especially in neuro informatics, which is my favourite study direction.

So here is the paper about fuzzy logic (pdf, german). It was my first attempt in using latex. Today, I love it.