Uncategorized

Machine Learning Invention Speaker Piston Cylinder – Where Every There Is Pressure To Convert Heat Into Work. I wonder if you can have a speaker with anti sound for ?higher efficiency. So you have a piston and a metal speaker in the cylinder. With iterated sounds for more work.So you ?can run a passenger plane with a small number of fuel piston-sound engines to electric motors

February 27, 2019 Per Lindholm

Uncategorized

Machine Learning – Select Reasons – A reason could be select elements in A that is equal to B. Closer to the truth. A=B. Another reason could be select elements which are close to a regression line of two or more model_i() outputs. So a reason could be an equation Ax=b.

February 27, 2019 Per Lindholm

Uncategorized

Machine Learning Physics Fusion Innovation Strategy – Are there co-inventions we can develop to get to fusion faster? Like getting over local minimas like with the car batteries and overhead wires?

February 26, 2019 Per Lindholm

Future generation starting at a higher level. Look back and picture beutify the process

Uncategorized

Machine Learning Test – Gray To Color. Cheat method. Probability more closer to the truth for small steps.

February 24, 2019 Per Lindholm

Not so big. But you can see colors emerging.

Uncategorized

Machine Learning Innovation – Cheat a little – You can learn by y = model(X) and update or you can occasionly cheat by y = np.random.rand() + yt.

February 24, 2019 Per Lindholm

every others simulated learning by a little cheating

Uncategorized

Machine Learning Math – Two ore more equal signs – x + y = x + y = x + y is True then there is a possibility for two or more equal signs. I think its just a cut off which x and y are possible // Per

February 24, 2019 Per Lindholm

to change the equation you probably need to change three sides of the equation

Uncategorized

Machine Learning Math – The difference between A and B if A == B is that the history is different. A might come from sum(vec) = A. Different future? Pi and probability belongs a little since the decimals of pi show random distribution parameters // Per

February 22, 2019 Per Lindholm

Uncategorized

Machine Learning Agriculture – Allow the plants to iterate?. Energy efficiency, (Th-Tc)/Th = n related to plant problems. Guess. If looks like the plant is enjoying themself maybe they are. Leaf shelter if too much sun == drapes?

February 22, 2019 Per Lindholm

Uncategorized

Machine Learning Guess – Over Problemsolvning – Label simulations that are related to the specific problem. Instead of direct transformation learning. Then ?fit the parameters of the simulation so it becomes more relevant to the problem.

February 22, 2019 Per Lindholm

Uncategorized

Machine Learning Math Guess – Invent any imaginary Problem and solve them and you get a more complete Mathematical Framework. For example. Math Forces are ?objects that move problems out of deep layer models()?

February 21, 2019 Per Lindholm

Uncategorized

Machine Learning Physics – Black Hole Network. When in doubt assume a network. Gives a black hole is a network that stops harmful things coming out. Black holes ?are not new physics but new intelligence due a lot of added problems.

February 21, 2019 Per Lindholm

Uncategorized

Machine Learning Math – Thought Math – Draw a bubble around a number or variable in an equation and imagine what this could mean. Then try to program solve a problem like this in machine learning.

February 21, 2019 Per Lindholm

Uncategorized

Machine Learning Paradigm Thought – If You Cant Split The Problem In Part Truths You Can ?Always Split It Into Many Guesses // Per Lindholm 2019-jan-27 Guess Mathematics

February 21, 2019 Per Lindholm

Uncategorized

Idea – Could a cause of a vulcano eruption be that not wanted material got under into the lava? CO2? So a the eruption has some meaning and is not just a physical effect.

February 21, 2019 Per Lindholm

Uncategorized

Machine Learning Idea – Using GAN asking to solve an equation. Showing x,y,z as the iterate towards the answer // Per

February 21, 2019 Per Lindholm

Uncategorized

Machine Learning Guess – As a possibility you can wait for some of equations in the system. Why solve all at the same time. Could produce more complex paths otherwise. Here I waited for y1-y2 -> 0

February 18, 2019 Per Lindholm

Uncategorized

Machine Learning Innovation Idea – Generalisation by same language model0(). y1 = model1(model0(X1)), y2 = model2(model0(X2)),y3 = model3(model0(X3)),… Where X1 = MNIST sample and X2 = Cifar10 sample And X3 = something else in image format. Image showing most linear recognition //Per

February 16, 2019 Per Lindholm

Uncategorized

Towards a Machine Learning Super Battery – Are different laser printed hole patterns different in their ion properties like capacity and transport? Laser aided CD printed like battery// Per

February 15, 2019 Per Lindholm

Uncategorized

Machine Learning Innovation – Air Condition Device – If There is a need of a compressor. I wonder if you can alter the surface function with a cylinder. Or better yet use a peltier element connected to the front end of the piston. So it cools the hot gas. //Per

February 14, 2019 Per Lindholm

Uncategorized

Blender 3D Innovation Idea – What if you take three render models() and iterate until you have enough intersecting pixel color values. Could this set be the inital value basis of a differential equation that solves the overall image with super high resolution? Venn Diagram Usage

February 14, 2019 Per Lindholm

Uncategorized

Towards A Super Battery – Inspired by ?saltpaper. Could sandpaper or paper with crystals be a no brainer thing of a super battery?

February 13, 2019 Per Lindholm

Uncategorized

Machine Learning Physics Idea – Predicting simulated turbulence with a w = model(rnge()).

February 13, 2019 Per Lindholm

Uncategorized

Machine Learning Math Idea – choice_division(a,b). Where you get a rest that moves between the even parts. Dynamic object?

February 13, 2019 Per Lindholm

Uncategorized

Machine Learning Physics – My most important tweet ever. Small fusion so that we have time and money to get the data samples. Million+?

February 10, 2019 Per Lindholm

Uncategorized

Machine Learning Complex Math – Doing some complex root finding with machine learning model() and chainer framework. y**2 = 1+1j gives y = 1.0986842+0.45508987j wow

February 10, 2019 Per Lindholm

Peroglyfer – Shorter than all geniuses put together.

Monthly Archives: February 2019

Machine Learning – Select Reasons – A reason could be select elements in A that is equal to B. Closer to the truth. A=B. Another reason could be select elements which are close to a regression line of two or more model_i() outputs. So a reason could be an equation Ax=b.

Machine Learning Physics Fusion Innovation Strategy – Are there co-inventions we can develop to get to fusion faster? Like getting over local minimas like with the car batteries and overhead wires?

Machine Learning Test – Gray To Color. Cheat method. Probability more closer to the truth for small steps.

Machine Learning Innovation – Cheat a little – You can learn by y = model(X) and update or you can occasionly cheat by y = np.random.rand() + yt.

Machine Learning Math – Two ore more equal signs – x + y = x + y = x + y is True then there is a possibility for two or more equal signs. I think its just a cut off which x and y are possible // Per

Machine Learning Math – The difference between A and B if A == B is that the history is different. A might come from sum(vec) = A. Different future? Pi and probability belongs a little since the decimals of pi show random distribution parameters // Per

Machine Learning Agriculture – Allow the plants to iterate?. Energy efficiency, (Th-Tc)/Th = n related to plant problems. Guess. If looks like the plant is enjoying themself maybe they are. Leaf shelter if too much sun == drapes?

Machine Learning Guess – Over Problemsolvning – Label simulations that are related to the specific problem. Instead of direct transformation learning. Then ?fit the parameters of the simulation so it becomes more relevant to the problem.

Machine Learning Math Guess – Invent any imaginary Problem and solve them and you get a more complete Mathematical Framework. For example. Math Forces are ?objects that move problems out of deep layer models()?

Machine Learning Physics – Black Hole Network. When in doubt assume a network. Gives a black hole is a network that stops harmful things coming out. Black holes ?are not new physics but new intelligence due a lot of added problems.

Machine Learning Math – Thought Math – Draw a bubble around a number or variable in an equation and imagine what this could mean. Then try to program solve a problem like this in machine learning.

Machine Learning Paradigm Thought – If You Cant Split The Problem In Part Truths You Can ?Always Split It Into Many Guesses // Per Lindholm 2019-jan-27 Guess Mathematics

Idea – Could a cause of a vulcano eruption be that not wanted material got under into the lava? CO2? So a the eruption has some meaning and is not just a physical effect.

Machine Learning Idea – Using GAN asking to solve an equation. Showing x,y,z as the iterate towards the answer // Per

Machine Learning Guess – As a possibility you can wait for some of equations in the system. Why solve all at the same time. Could produce more complex paths otherwise. Here I waited for y1-y2 -> 0

Machine Learning Innovation Idea – Generalisation by same language model0(). y1 = model1(model0(X1)), y2 = model2(model0(X2)),y3 = model3(model0(X3)),… Where X1 = MNIST sample and X2 = Cifar10 sample And X3 = something else in image format. Image showing most linear recognition //Per

Towards a Machine Learning Super Battery – Are different laser printed hole patterns different in their ion properties like capacity and transport? Laser aided CD printed like battery// Per

Machine Learning Innovation – Air Condition Device – If There is a need of a compressor. I wonder if you can alter the surface function with a cylinder. Or better yet use a peltier element connected to the front end of the piston. So it cools the hot gas. //Per

Blender 3D Innovation Idea – What if you take three render models() and iterate until you have enough intersecting pixel color values. Could this set be the inital value basis of a differential equation that solves the overall image with super high resolution? Venn Diagram Usage

Towards A Super Battery – Inspired by ?saltpaper. Could sandpaper or paper with crystals be a no brainer thing of a super battery?

Machine Learning Physics Idea – Predicting simulated turbulence with a w = model(rnge()).

Machine Learning Math Idea – choice_division(a,b). Where you get a rest that moves between the even parts. Dynamic object?

Machine Learning Physics – My most important tweet ever. Small fusion so that we have time and money to get the data samples. Million+?

Machine Learning Complex Math – Doing some complex root finding with machine learning model() and chainer framework. y**2 = 1+1j gives y = 1.0986842+0.45508987j wow