random_state, please help

Question

I need to understand what exactly the random_state parameter does in the model
from my understanding its just a value that i should keep it fixed in the train_test_split and in the model object so it does not give me a random accuracy score everytime.

so my question is everytime i change the random state, my model accuracy gets better or worst, so i decided to put my model in a loop to find a random_state value that gives the highest accuracy to my model, is that a right thing to do or not

for example:

scores = []

for n in range(1, 201):

X_train, X_test, y_train, y_test = train_test_split(X, y, train_size=0.8, test_size=0.2, random_state=n)

model = MLPClassifier(random_state=n)

model.fit(X_train, y_train)

scores.append(model.score(X_test, y_test))
print(scores)

Answer 1

This will help you:

https://medium.com/analytics-vidhya/what-is-random-state-in-machine-learning-84c1c9dffaad