Vishal Mehta vishalmehta1991

|  ######cuML######                 |  ######Sklearn######                 |
|                                   |                                      |
|  from cuml import                 |  from sklearn.ensemble import        |
|  RandomForestClassifier as cuRF   |  RandomForestClassifier as sklRF     |
|                                   |  import multiprocessing as mp        |
|                                   |                                      |
|  # cuml Random Forest params      |  #sklearn Random Forest params       |
|  cu_rf_params = {                 |  skl_rf_params = {                   |
|     ‘n_estimators’: 25,           |      ‘n_estimators’: 25,             |

Instance	Red	Green	Blue	Size (cm)	Fruit (Label)
1	0.0	1.0	0.0	20	Water Melon
0	1.0	0.0	0.0	7.0	Apple
5	1.0	0.0	0.0	0.8	Cherry
2	1.0	0.0	0.0	1.0	Cherry

Instance	Red	Green	Blue	Size (cm)	Fruit (Label)
4	1.0	0.0	0.0	1.0	Strawberry
4	1.0	0.0	0.0	1.0	Strawberry
1	0.0	1.0	0.0	20	Water Melon
3	0.0	1.0	0.0	7.5	Apple

Instance	Red	Green	Blue	Size (cm)	Fruit (Label)
5	1.0	0.0	0.0	0.8	Cherry
0	1.0	0.0	0.0	7.0	Apple
0	1.0	0.0	0.0	7.0	Apple
4	1.0	0.0	0.0	1.0	Strawberry

Instance	Red	Green	Blue	Size (cm)	Fruit (Label)
0	1.0	0.0	0.0	7.0	Apple
1	0.0	1.0	0.0	20	Water Melon
2	1.0	0.0	0.0	1.0	Cherry
3	0.0	1.0	0.0	7.5	Apple
4	1.0	0.0	0.0	1.0	Strawberry
5	1.0	0.0	0.0	0.8	Cherry

	from cuml.dask.ensemble import RandomForestClassifier as cuRF_mg
	# cuml Random Forest params
	cu_rf_params = {
	‘n_estimators’: 25,
	‘max_depth’: 13,
	‘n_bins’: 15,
	‘n_streams’: 8
	}

	# Start by setting up the CUDA cluster on the local host

	(A) Initialize a bit mask indicating which samples are contained in each node
	(B) Initialize a “node map” indicating which nodes are present at each level
	(C) ForEach(tree_level)
	1. Find the node id of all data samples, using the bit mask
	2. Compute the possible splits for all bins, all columns and all nodes
	3. Find the best split for each node
	4. Update the bit mask and sparse node map to feed the next level