Hi guys, is there anyone who works with R, can tell me what is the best method to obtain an adjusted model ?
hello too, do you mean a tuned model?
Yes, i mean is it better to work with neural network or decision tree or other, and what can i do with features like data processing to improve my model?
the type of model to use highly depends on the data you are using to train your model.some data will fit well with neural nets, random forest etc and some will fit well with knn, svm etc.
model accuracy also depends on various factors such as treating missing values and outliers, feature selection and engineering, cross validation and many others. although, in many cases ensemble learning models mostly have a higher acuraccy but i would still advice you to try out various models and then choose which fits your data well
thanks a lot, it's very clear, i've used nnet gave me 88% accuracy after ross validation.