# Discuss kNN, test/train, random projection, unit vector, vectored matrix, hamming distance, stft, Y=(aAX1:M)

Suggestion Investigation

Looking for suggestions or guide for how to setup a clean approach and discussion on how to apply a python suggested way to solve this challenge

Looking for suggestions or guide on how to create a random project matrix and represent this as vectorized N (e.g., 3,4,5,6) rows of the magnitude STFT

Looking for suggestions or guide on how to create random projection following Y = sign (AX1:M,:)

Problem Effort

Working on .mat file (training, test) samples. Trying to understand how to apply custom kNN classification (from scratch) on the binary data test/train.

Need to test and compare the bit strings on the training data with Hamming distance.

Need to create a random project matrix to compute the number of nodes of dimensions of sample (train/test) and represent this a vectorized 3,4,5,6 rows of the magnitude STFT, making X and 255 Z for a new dimension random projects.

A random project should follow Y = sign (AX1:M,:). This matrix with random values must be row vectors, or projection vectors and ensure unit vectors ∥Ai,:∥2 = 1.

On the kNN classification, every sample (test example) is to find the K nearest neighbors out of 112 training samples.

Outcome Results

random project matrix : Y = sign (AX1:M,:)

unit vectors must be ∥Ai,:∥2 = 1

vectorized matrix with N rows of the magnitude STFT