klog

yes. another resolutions page. I’ll call them plans, though, as “resolutions” sounds a bit certain. in my experience, nothing is certain, even promises.

so, my plans:

• lose a half-stone. when I married Herself, I was 12 stone. this was probably partly caused by the anti-depressants I was on at the time. since then, it took me a few months to manage to cut down to 11 stone. my ideal weight is more like 10 stone, but I’m finding it incredibly difficult to do that - I just seem to be stuck at 11. so, this year’s plan is to get down to 10.5.
• build the new robot - I want a very small form-factor machine (gumstix or linutop). it should have caterpillar tracks, a robotic arm, and a cutting tool. my current bot is about 9′x14′. the new one should be at most 6′x6′. the software is not yet complete, but at least this will get the hardware out of the way
• get my finances back under control. for the last few years, I’ve had negative values in almost all of my bank accounts. about time I managed to fix that. I have one account under control by basically denying myself access to it and setting up a standing order into it. the same plan should work with everything else. the plan is to on average raise the total value of each account by a set value every month. this value (none of your damned business ) should be measured on a specific day each month.
• build a collapsible mini-ramp. I can’t make it to the local skate-park very often, so the plan is to build something that I can wheel out to the garden when I feel like some exercise. it should be big enough that I can use my bike on it as well as my skateboard.

notice that none of these are vague. it’s been shown that when a resolution is vague, it usually does not happen. an exact plan, though, gives solid goals that can be reached.

Last week, I wrote a neural network that could balance a stick. That was a simple problem which really only takes a single neuron to figure out.

This week, I promised to write a net which could learn to recognise letters.

demo

For this, I enhanced the network a bit. I added a more sensible weight-correction algorithm, and separated the code (ANN code).

I was considering whether hidden inputs were required at all for this task. I didn’t think so - in my opinion, recognising the letter ‘I’ for example, should depend on some information such as “does it look like a J, and not like an M?” - in other words, recognising a letter depends on how confident you are about whether other values are right or wrong.

The network I chose to implement is, I think, called a “simple recurrent network” with stochastic elements. This means that every neuron reads from every other neuron and not itself, and corrections are not exact - there is a small element of randomness or “noise” in there.

The popular choice for this kind of test is a feed-forward network, which is trained through back-propagation. That network requires hidden units, and each output (is it N, it it Q) is totally ignorant of the other outputs, which I think is a detriment.

My network variant has just finished a training run after 44 training cycles. That is proof that the simple recurrent network can learn to recognise simple letters without relying on hidden units.

Another interesting thing about the method I used is how the training works. Instead of throwing a huge list of tests at the network, I have 26 tests, but only a set number of them are run in each cycle depending on how many were gotten right up until then. For example, a training cycle with 13 tests will only be allowed if the network previously successfully passed 12 tests.

There are still a few small details I’d want to be sure about before pronouncing this test an absolute success, but I’m very happy with it.

Next week, I hope to have this demo re-written in Java, and a new demo recognising flowers in full-colour pictures (stretching myself, maybe…).

As always, this has the end goal of being inserted in a tiny robot which will then do my gardening for me. Not a mad idea, I think you’re beginning to see - just a lot of work.

update As I thought, there were some points which were not quiet perfect. There was a part of the algorithm which would artifically boost the success of the net. With those deficiencies corrected, it takes over 500 cycles to get to 6 correct letters. I think this can be improved… (moments later - now only takes 150+ to reach 6 letters)

robots!

So anyway, I moved house (long story short), meaning that I get to think more clearly, as the house is less cluttered, and the route to work involves crossing less roads.

This morning, I was thinking about my current project - I’m writing a recurrent connectionist network so my new robot can learn to recognise things like grass and rubbish (to cut the former, and remove the latter).

The walk was getting tiring, so I was thinking about segways as well, and wondering how easy it might be to make one.

This eventually evolved into an idea for a new transport system - you get a load of little robots (my gardening ones, for example), and get them to form a platform. Then a load more of them form another platform on top. Then, you stand on the top.

The “carpet” would move in the direction you lean. Of course, the speed wouldn’t be too impressive, but it would be better than walking.

When the lower layer encounters a rock on the road or something, it moves around it. The upper layer robots interlock with each other to allow the lower level bots to do this without having too much pressure from above.

When you reach where you are going, the robots then disperse and continue their gardening around the new area.

You could even form a baggage train using this idea - a few carpet networks would follow each other in marching-ant form.

This would be easier to do than to create a robot which does your gardening for you…