Particle Swarm Optimisation

Поділитися
Вставка
  • Опубліковано 1 гру 2024

КОМЕНТАРІ • 7

  • @habtelejebo895
    @habtelejebo895 2 роки тому +1

    Love it! Thank you...can you do a video on how to train neural network using PSO?

    • @enlightenment609
      @enlightenment609  2 роки тому +2

      Its actually not that difficult. PSO/GA etc can be used to tune the hyper parameters of a Neural Network; the training itself will be internal to the NN. To do this, you simply plug in the NN into the fitness function of a PSO and then initialise it with the PSO generated values of the hyper parameters.

  • @dennisbarrios7562
    @dennisbarrios7562 2 роки тому +1

    Hello, Thank you... you have a nice video presentation. I have a question, is it possible to identify the boundaries in PSO?

    • @enlightenment609
      @enlightenment609  2 роки тому +1

      Hi Denis. It depends what you mean by "boundaries"? If you mean boundaries of the search space, then the user (problem provider) defines them. If it something else, please explain.

    • @dennisbarrios7562
      @dennisbarrios7562 2 роки тому

      @@enlightenment609 Yes, I mean the boundaries of the search space, in the equation I cannot determine what is the boundaries of the search space. If none, Can you please help me identify where to put in the equation or in the algorithm? Thank you in advance.

    • @enlightenment609
      @enlightenment609  2 роки тому

      @@dennisbarrios7562 those boundaries are dettermined by the user. For example, suppose you are optimising some pressure values. Then you would know how far up or down the pressure can be turned. So the user knowledge determines those boundaries. Similarly, say you want to set working hours then you would know what is the minimum and maximum working hours. So you can think further examples like that.

  • @kittycatgirl1139
    @kittycatgirl1139 Рік тому

    How does it work with laser pm 2.5 particle sensors ? What’s the objective