315 - Optimization using Genetic Algorithm
ฝัง
- เผยแพร่เมื่อ 11 ก.ย. 2024
- Code generated in the video can be downloaded from here:
github.com/bns...
The genetic algorithm is a stochastic method for function optimization inspired by the process of natural evolution - select parents to create children using the crossover and mutation processes.
Coding it in python: The algorithm consists of the following key steps:
Initialize a population of binary bitstrings with random values.
Decode the binary bitstrings into numerical values and evaluate the fitness (the objective function) for each individual in the population.
Select the best individuals from the population using tournament selection based on the fitness scores.
Create new offsprings from the selected individuals using the crossover operation.
Apply the mutation operation on the offsprings to maintain diversity in the population.
Repeat steps 2 to 5 until a stopping criterion is met.
Hi, I see the channel approaching 100K Subscribers, wish you all the best.
Thank you for your great effort.
Thanks 😊
A most awaited video sir.... Thank you so much....can you please upload a seperate video for hyper parameter tunning using genetic or merahuristic algorithms for image classification tasks??? Sir
Hello
Dear Dr. DigitalSreeni,
I adore all of your video courses. Keep going! However, I have a question:
you put this parameters scientifically or
# Define the algorithm parameters
algorithm_param = {'max_num_iteration': 10,\
'population_size':100,\
'mutation_probability':0.01,\
'elit_ratio': 0.01,\
'crossover_probability': 0.9,\
'parents_portion': 0.3,\
'crossover_type':'uniform',\
'max_iteration_without_improv':None}
this one Just testing parameters ...?
If I understood correctly, the genetic algorithm in this case was not used to train the model itself, but to optimize the best input values from the data, correct? It may be interesting to see a genetic algorithm do the process of training of the network itself. Instead of the common model.fit function. I mean considering each individual as an independent set of weights, and to optimize the weights and biases values. So basically: Set the weights of the model --> predicting --> evaluating the fitness depending on the prediction output --> (repeat steps for N population elements) --> selecting fittest --> crossover --> mutation --> etc...
Hey, thank you so much for such a great video!
A question: How to include constraints in this implementation?
Once again you have been a great help/
Thank you this informative video.
I have a doubt that if I have trained and ANN model instead of RandomForest regressor then also I have to train "X" and "y" and then, how to define e new model with ANN model?
@DigitalSreeni Sorry, I did not understand the purpose of using genetic algorithm in this video. Can you please explain? Thanks.
Great content! Is there a way to access the CSV file you are using?
Already I saw a video for the convertion of ML model into web application using "csv" dataset, which was already in the channel. But please upload a seperate video for image classification tasks. If possible means take an example model as "vgg 16 as a feature exractor" which is already in your channel sir.. please.... waiting 🤞
Hi! Can you upload a video with an example of img2img diffusion model? (DDIM for example, for denoise image, segmentation or superresolution), would help me a lot.
Thank you so much
One more request sir.... please upload a seperate video for "how to convert the DL or ML model (image classification tasks) into web application or Android application using flasks...."?
Awesome video ❤
Glad you liked it
Thanks!
Thank you
Genetic Information is the TimeStamps embedded in the Atoms. Life gets its energy from the time differences of the biological elements in our body. I am able to explain every physical, chemical and biological processes using this discovery.