Okay How can I do this model evaluation it is because you are using this Java. So, he will be around the same as this classification. So, I copy the whole method okay. I paste it here then I will change this to evaluation model evaluation. So I change we evaluate our model trains, Tessa algorithm train test testing test all will be the same. So, the Import Data Source and bodies are the same if algorithm equal to linear linear regression equal new linear regressions, classifier c is equal linear and then C is your classifier we will change something here.

So, here are classified instances. So for each row in testing South Asia instances are each observation in the testing set. we classify the instance. So to evaluate model, we can do something like this evaluation or evaluation. Evaluate equal new evaluation using trainset. Okay, evaluate dot Eva, da, Mada CLS test sec.

Okay. Then we can see dot out. dot brain ly Okay evaluate to some Reese pre k slash, so this is our linear so linear regression okay then resolve slash n okay then slash n okay to summary resolve Okay, here should be false okay we can use this method to print a print result here so I can remove this one yeah I have to employ class here. Okay valet model a valet more should be more system error no System dot out dot println I print the result. Okay, so easy linear regression I'm not going to evaluate a linear regression I have to change because two 3.6 all these. So, I will do the same for the Allah algorithm.

So, what we change now is our evaluation until these are System dot out dot print line. So, I will do something like this, it should be after your classifier will change something IDs ah tree so, these are linear regression of a change to decision tree okay then I will replace here So, K a near regressor I change to k n k then for Naive Bayes replace it also. Okay. So here I remove and I changed to Naive Bayes and for MLP in summary here I changed this to MLP Amati the perceptron okay then I can call the evaluate our model. So I can do something like this okay. Evaluate a model that I can remove this classification I should have some error sama Because, in this evaluate our model, when we evaluate a model, we need a test set with labor.

So for our Iris data set, we don't have the labor. So to evaluate our model, we need a ERISA test to add labor here. Why do we need this labor? So, let's say when we are when we evaluate evaluate a model, so, after we train the model V, these are Iris data set, then we put in this testing set into the model and then this model will predict the cross here. Then we will based on the cross that we predicted and these are across here. So we will do the Evaluation we do all the calculation for the accuracy.

So, we will do the equation or calculate the accuracy calculate the accuracy using all those confusion matrix and so on. So, in confusion matrix we had extra So, extra will be the actual labor here and the predictor our labor we will compare our actual and predicted labor and then we will get our accuracy, precision and recall from this confusion matrix Okay, so, for our case I can just use the same data I should have some other accuracy here. Okay, so MLP reside we'll have all the accuracy here. If I change to other algorithm, so let's say Naive Bayes, so I saw some result here. So okay.