Machine learning question, Solve this problem using KNN method?

Here is the link to the website so you could access the question along with the links for the data sets

https://grid.cs.gsu.edu/zcai/course/4980-6980/

Classification: Classification is to identify which category a new observation belongs to, on the basis of a training dataset. There are five datasets. For each dataset, we provide the training dataset, training label, and test dataset. Please use the training dataset and training label to build your classifier and predict the test label. A class label is represented by an integer. For example, in the Ist dataset, there are 4 classes where 1 represents the 1st class, 2 represents the 2nd class, etc. Note that, there exist some missing values in some of the dataset (a missing entry is filled by 1.00000000000000e+99), please fill the missing values before performm your classification algorithm. TrainData 1 contains 150 samples with 3312 features. Testdatal contains 53 samples with 3312 features. There are 5 classes in thi dataset. TrainData 2 contains 325 samples with 28 features. Testdata2 contains 198 samples with 28 samples. There are 4 classes in this dataset. TrainData 3 contains 6300 samples with 13 features. Testdata3 contains 2693 samples with 13 features. There are 9 classes in this dataset. TrainData 4 contains 2547 samples with 112 features. Testdata4 contains 1092 samples with 112 features. There are 9 classes in this dataset. TrainData 5 contains 1119 samples with 11 features. Testdata5 contains 480 samples with 11 features. There are 11 classes in this dataset. TrainData 6 contains 612 samples with 142 features. Testdata5 contains 262 samples with 142 features. This is not a cl problem. You are asked to predict the real value. (Graduate Students Only) Classification: Classification is to identify which category a new observation belongs to, on the basis of a training dataset. There are five datasets. For each dataset, we provide the training dataset, training label, and test dataset. Please use the training dataset and training label to build your classifier and predict the test label. A class label is represented by an integer. For example, in the Ist dataset, there are 4 classes where 1 represents the 1st class, 2 represents the 2nd class, etc. Note that, there exist some missing values in some of the dataset (a missing entry is filled by 1.00000000000000e+99), please fill the missing values before performm your classification algorithm. TrainData 1 contains 150 samples with 3312 features. Testdatal contains 53 samples with 3312 features. There are 5 classes in thi dataset. TrainData 2 contains 325 samples with 28 features. Testdata2 contains 198 samples with 28 samples. There are 4 classes in this dataset. TrainData 3 contains 6300 samples with 13 features. Testdata3 contains 2693 samples with 13 features. There are 9 classes in this dataset. TrainData 4 contains 2547 samples with 112 features. Testdata4 contains 1092 samples with 112 features. There are 9 classes in this dataset. TrainData 5 contains 1119 samples with 11 features. Testdata5 contains 480 samples with 11 features. There are 11 classes in this dataset. TrainData 6 contains 612 samples with 142 features. Testdata5 contains 262 samples with 142 features. This is not a cl problem. You are asked to predict the real value. (Graduate Students Only)