Make n_estimators in Isolation Forest a configurable option

[stevenlujpl]

No description provided.

[stevenlujpl]

I've updated the iforest_outlier_detection.py script to make n_estimators a configurable option. To use iforest ranking method, we will need to provide the n_trees parameter in the config file. For example,

outlier_detection: {
    iforest: {
        n_trees: 100
    }
}

[stevenlujpl]

@urebbapr You can use the following API to call isolation forest for the sequential isolation forest implementation. Please let me know if there is any question. Thanks.

dora/dora_exp_pipeline/iforest_outlier_detection.py

Lines 11 to 12 in 3410aa5

	def _rank_internal(self, data_to_fit, data_to_score, data_to_score_ids,
	top_n, seed, n_trees):