knn | A k-nearest neighboor classifier algorithm | Learning library

I want to save a Tensorflow model and then later use it for deployment purposes. I dont want to use model.save() to save it because my purpose is to somehow 'pickle' it and use it in a different system where tensorflow is not installed, like:

I am currently using the scikit learn module in order to help with a crime prediction problem. I am having an issue batch coding the entire Dataframe that I have with the knn.predict method.

How can I batch code the entire two columns of my Dataframe with the knn.predict() method in order to store in another Dataframe the output?

I would like to train a SVM and KNN model, I was wondering if anyone could advise for training a Machine Learning model which processing is more appropriate?

I have 3 columns namely Models(should be taken as index), Accuracy without normalization, Accuracy with normalization (zscore, minmax, maxabs, robust) and these are required to be created as:

I want to impute missing values by KNN, and I use this method to select best K:

I wanted to use the optimal k value to conduct kNN clasification and predict the dependent variable diabetes in test set using train set and compare the results with the real values.

I've already got optimal k value and got the accruacy already. After that, I wanted to compare the results with the real value with using confussionMatrix but I got the problem with the different length.

I've already checked that nrow and length quantities are same(with 74) but it still have same problem.

Could you help me to overcome this problem?

My codes are as like below

I just wanted to conduct a kNN classification with the situation when k is 3. I would like to predict the dependent variable “diabetes” in valid set using train set and calculate the accuracy.

But I faced to the error message with

Error in knn(train = TrainXNormDF, test = ValidXNormDF, cl = MLdata2[, : 'train' and 'class' have different lengths

I can't solve this problem with get approach with

background: I have a machine learning model in which given an object returns an embedding vector with dimension d, the model is trained in a way such that the semantic similarity of two embedding vectors is very close. Now, the verification process is relatively simple, I can take something like the cosine similarity of the two vectors. For recognition, it's a little bit complicated, either I can loop through all the anchor documents and compare the cosine similarity, or use something like kNN (online).

problem: I have a list of embedding vectors, each vector has a dimension d, with length N. Each vector contains floating-point data.

What will be an efficient data structure + algorithm that can do the following:

1. Can add a new vector with a unique ID to the list efficiently (<= logarithmic complexity)
2. Search with a random vector in the list, and retrieve top k vectors, such that the Manhattan distance / L1 norm is minimum for those vectors efficiently (hopefully, <= logarithmic complexity).

example:

1. I use Python.Net for C# interaction with Python libraries. I solve the problem of text classification. I use FastText to index and get the vector, as well as Sklearn to train the classifier (Knn).During the implementation, I encountered a lot of problems, but all were solved, with the exception of one. After receiving the vectors of the texts on which I train Knn, I save them to a separate text file and then, if necessary, use it.