[/b/] [/d/] [/tu/] [/a/] [/ph/] [/wa/] [/cg/] [/t/] [/p/]
Stochastic Pattern Recognition Dramatically Outperforms Conventional Techniques>A stochastic computer, designed to help an autonomous vehicle navigate, outperforms a conventional computer by three orders of magnitudehttp://www.technologyreview.com/blog/arxiv/27598/?p1=blogsЧто думаешь, анон?
Stochastic Pattern Recognition Dramatically Outperforms Conventional Techniques
>A stochastic computer, designed to help an autonomous vehicle navigate, outperforms a conventional computer by three orders of magnitude
http://www.technologyreview.com/blog/arxiv/27598/?p1=blogs
Что думаешь, анон?
Как оно работает, ОП?
>>76420>Stochastic Pattern Recognition Dramatically Outperforms Conventional Techniques>Что думаешь, анон?какая неожиданность! журналешлюхи уже совсем пизданулись.
>>76420
>Stochastic Pattern Recognition Dramatically Outperforms Conventional Techniques>Что думаешь, анон?
какая неожиданность! журналешлюхи уже совсем пизданулись.
>>76422Если я правильно понял статью, примерно как из анекдота про логарифмическую линейку. "Что-то около 30-и".Мимо крокодил
Интересно, для каких классов задач эта штука пригодна.Пример там какой-то искусственный.>We verified the proposed patter-recognition technique to orient an autonomous vehicle in aknown environment. At each time step, the vehicle computes the distance to the nearestwalls (parameters n,s,e and w, in Fig. 8). These parameters are the inputs of seven MSCs(one for each zone of the plane) configured with specifics mean and sigma values(parameters η and N in Fig. 4). Once the vehicle recognizes the environment (the classrecognized is the zone where it is the vehicle), a predefined movement is selected with theobjective to reach zone A (see Fig. 8). In different experiments were the vehicle was placedrandomly in the plane, the time steps needed to reach zone A were computed. In all cases,the vehicle was able to reach zone A with a nearly minimum trajectory. In Fig. 9 wecompare the time steps required to reach zone A (symbols) with the optimum number oftime steps (solid line). Results show that the itinerary followed by the vehicle is very closeto the optimum trajectory thus demonstrating that the stochastic system is able to recognizethe zones at each point of the path followed.
Интересно, для каких классов задач эта штука пригодна.Пример там какой-то искусственный.
>We verified the proposed patter-recognition technique to orient an autonomous vehicle in a
known environment. At each time step, the vehicle computes the distance to the nearestwalls (parameters n,s,e and w, in Fig. 8). These parameters are the inputs of seven MSCs(one for each zone of the plane) configured with specifics mean and sigma values(parameters η and N in Fig. 4). Once the vehicle recognizes the environment (the classrecognized is the zone where it is the vehicle), a predefined movement is selected with theobjective to reach zone A (see Fig. 8). In different experiments were the vehicle was placedrandomly in the plane, the time steps needed to reach zone A were computed. In all cases,the vehicle was able to reach zone A with a nearly minimum trajectory. In Fig. 9 wecompare the time steps required to reach zone A (symbols) with the optimum number oftime steps (solid line). Results show that the itinerary followed by the vehicle is very closeto the optimum trajectory thus demonstrating that the stochastic system is able to recognizethe zones at each point of the path followed.
Да вы лентяи же! И оп - самый ленивый.
- wakaba 3.0.7 + futaba + futallaby -