I wish everyone a merry Christmas
Machine learning as it is today is pretty simple. You have input, a and output b with a function net(a) in between.
Doesn’t this resemble old math. You have one object a that is proportional to an object b.
When does this not hold. If net(a) is singular for some a = a(k). But how could net(a(k)) get singular. One feature of 1/x is perhaps the closeness to a possible singularity at x = 0.
So I thought. If you have two networks. A rational object if you will. That is. net(a) = net0(a)/net1(a). This might get me new possibilities. Here I will train net0 and net1 as if they were separate neural networks. That is. Some hidden layer networks.
So the idea was to include a feature of math. The closeness to a singularity.
Maybe you need to use network objects obj1, obj2, obj3,… as in ordinary math and just train them. Could there exist like a reduced taylor series for network objects?
So here you can create a built in gain function in the model.
Just for inspiration. A quick battery idea just as a philosophical possibility.
So my idea is to capture the energy from electrons while they are flying at high speeds.
I think its possible to create such a battery with electrons in vacuum with high voltages. However vacuum might not be so ideal since it does not provide an adaptable function. So maybe the battery would use a neon tube with ?coils.
Maybe the battery today is a heavy bottle neck and you need an electron ?multiplier that does not take so much weight.
If nothing else I had fun making the rendered image.Maybe an extreme battery should have extreme properties like high electron ?speed. The big C battery.
I wonder. Why don’t we have a City Mission For Homeless which collects donated wood planks and timber. To be used for building tiny house villages for the homeless.
For Sweden who has yet to pick up this housing first strategy. I wonder if one idea could be to trade wood planks for helping build these small houses.
I mean. Wood is an expensive product and labor costs could also be an obstacle. So when a home carpenter builds a porch. He or she could perhaps help build a tiny house. For some amount of wood.
Often there are 3 for 2 for the same garment. This is that smart. What I think could work is a 2 for the price of 1 deal. For woman and mens clothing combined.
This way ?you will feel happy when shopping for clothes. Something for her and him.
The assumption is that the universe has calculated ?everything. Then there should exist a relation between x,y,z position and time for any small object.
So for the smallest object I assume there is some kind of a network with x,y,z,time as input data. Just because the calculation exists as a possibility.
Then I guess there is network answer that is fulfilled for x,y,z,t in the network. If the network just have one node in the end the answer is single number then maybe its a constant for every such object.
A test I made. Aiming for something constant. Be it a constant vector or number. I really got to many candidates for constants.
Since Israel and Palestine has shown restraint. I wonder if European countries can defuse this situation with a trust seeking declaration on Jerusalem.
One idea is to have two embassies each. One in Tel Aviv and one in Jerusalem. (??)
Some sort of trust seeking declaration on Jerusalem from the European countries to Israel is needed.
The idea is simple.
I wonder. For fusion. Could ultra determinism help. By this I mean. There could exist a wise relationship for fusion technology. Since it might be dangerous. That is. You need 100% knowledge when and where fusion takes place.
By this relationship. A large hot volume for fusion like the tokamak is not the start of this technology. I guess a smaller device which is optimized for answering the two questions. Where and when exactly fusion takes place. Is the way to go.
I made a test of an error recognition with the irids dataset. The idea was to add an additional column to the target examples. The column was just a 1 for true examples and 0 for random generated examples.
So what I did first was to train the model network on the just the true examples.
Then I added double the amount of randomly generated data in the input data. With that random data I run it through the model. Predicting which class the random data would belong to. This falsie predicted data I could then use as false examples. That is. I added a 0 column to that target data.
This data together with the old data got me new data to train my second model. One which could tell if the input was ?wrong and still predict the correct output.
Although it took longer to converge.
So with this test I could tell if the sample looked like it came from a random generator giving close to 0.0 at the first column. Similar it could tell a value of .99 for data coming from true examples.
Otherwise the network would just use the method of elimination which is not so good.
Code is just for me to remember the idea. Not anything to copy. Maybe to correct though : )
For students to feel secure in school. Should there not exist a school-peace fund who comes up with anti bullying projects and ideas.
Somehow you need a broader approach to reduce tragedies like school shootings. A peaceful school where everyone can feel safe and happy. I don’t think its that hard. From my point of view there needs to be different projects running all the time. This way school children know their well being matters.
For important train commuters lines I believe something has to be done. A failure should not affect the peoples whole day. Trying to find a bus, taxi or getting stuck at the Central City Station.
Its not enough to build additional tracks for higher traffic volume. The tracks are dependent on the same failures. If built like usual.
The idea is to build an additional independent train commuter line. Such that if one line gets an error people can go to the other line.
So for example. A two independent line system should have independent power supplies and maybe a big and a little Central Station.
With this idea. People get a much needed backup and for a carbon emission free future public transport must have an excellent uptime with zero stops.
One idea is that to create extra stations along the extra way. Since stations are gold to the neighborhood. The number of housings will ?increase overall. If you don’t make the train system error resilient then the problem with going by train will persist.
A quick idea
Inspired by the digital age. Why not build an identical copy of Jerusalem. Let the Palestinians decide if it will be their capital.
The reason for this is pretty simple. If you build so that you have two or three Jerusalem there should be less probability for arguing. There is a Jerusalem for Israel and for one Palestine and possibly one left as a symbol for peace.
Since we have large TV with high resolutions. I thought why not add a digital border and scale down the TV image a little.
Then the border could have different themes be it new snow, advent and Christmas.
This would not be a problem for a smart TV dongle with a TV input signal. Just an image overlay.
Could we improve the chances of peace with food? Or rather. It could be that perhaps the wrong type of sugar or too much sugar for longer periods of time influences us in a bad way.
I would argue that certain spices in food could have a positive effect on the brain. I base this assumption on the feeling I get when eating curry chicken. Thats my guess anyway.
It would be better for Israel if they respectfully declined the American decision for Jerusalem. For the time being at least. Considering that it could potentially cause too much harm.
Its not necessary. Nothing says you cant get a second chance at this even after a retraction. I believe peace and trust can be built this way.
Maybe we could have three cities and Jerusalem being culture protected by two countries law.
Innovating peace as fast as I can.
Could every country have a minister of peace. Also. Could the bureaucracy be adapted towards peace. I think this is a sustainable way forward and a modern approach.
A quick idea.
Peace is a complex problem. So it will require something from all us. To have a sustainable peace I have a feeling it will cost a lot of money.
Therefor I wonder if this peace capacity could be financed by a tax or a fee.
Since so many people will be involved I wonder if the money could stay in the persons bank account as seperate account. I feel this is the safest solution. So when money is required it can be reserved. Not played with by some hotshot investor.
The idea is simple.
Two cities are better than one because when you select a capital there is a choice inbetween.
A quick math idea this morning.
I argue that our problem descriptions in probability are often wrong. They state the probability as a non complex value when it ?is a machine learning network. The reason is that a network can ?assume any function.
So what you need is data. Then feed that initial data to a network model to converge.
You can then test the network for probability calculations. However. Since different models can converge to the same data. There should exist a way to estimate this uncertainty.
So in everything you should ask yourself. Could networks assume these problems. If so. Then try it.
The idea is simple.
The world needs peace to survive the times to come. I would argue that you can turn this apparent difficulty to your own advantage. Its very smart to build a new capital. Somewhere else than in the disputed Jerusalem. I mean these things cost a lot money. Money that will be spent to ensure a prosperous future for the younger generation to enjoy.
I believe a ?global bank can be established to finance a project like this.
Come to think of it. Would the city need to be built with lots of tech and other innovations to make it resilient to climate change for instance. I like this idea. What if you have the best architects, engineers and planners to figure this out.
A quick language learning idea.
I grew up with humorous adventure games like Monkey Island, Simon the sorcerer, Indiana Jones and more. They were all in English so in school the teacher had an easy time teaching this language.
So if you are in the process of learning another language like French, German, Arabic or English. Why not change the language of the game or the software your are using.
An easy way is to manage the phone/tablet/computer in another language. Maybe you can have your old phone in the language you want to learn and your current phone to Google the words.
Philosophically you have the option to think on your own or in a group.
I would argue that the best solution to a complex problem of any kind is a network. Since it can assume any options or option variant.
For example it might turn out that you get the best results if you generate ideas first on your own and then after some time you meet to discuss different ideas.
In my previous idea that the complexity problem is solvable with a network. I connected the weights in a continuous fashion to get a clearer picture. The smoothed image of the weights matrices.
From this I realized that you can connect the weights in a machine learning neural network. This so they are not so independent. Not looking so random when combined. So this gives me the idea that the weights are complex objects. Which are solvable each with a network.
So what input does the weight network have. Here I take inspiration from the universe. Molecules provide a clue. Here you have connected weights through bonds. But molecules are not so big.
Then this idea explains why we have gravity. Its just input data from all other weights to the particular weight. For the weights to solve more problems they need more information.
That is. Molecular forces provide input from the immediate surrounding weights and gravity provides information from all the rest of the weights.
So to sum this idea. Everything that seems complex is just networks. // Per Lindholm
Come to think of it. If energy runs the network it could resemble the error in a machine learning network. From that I realize that minimizing the error could happen in more ways. ?Either you disperse the error so it get evenly distributed or you shift the error to another place in the network. So at least locally you get a low amount of error. This with a limited amount of weights.