Solution Overview 

Solution 1: Weighted NSForest

• Implement a weighting scheme using a pairwise distance matrix, emphasizing features from closely related clusters and penalizing features from distant clusters

Solution 2: Elimination of Cluster Biases

• Run NSForest, but downsampling large clusters to thresholds of 50, 100, 150, 200, and 500 cells

• Evaluate resulting NSForest performance using the F-beta evaluation metric