Remove broken algorithm option

[jeremykubica]

The code for gradient_boosted_trees() is commented out with a note "we need to find a non-bugged version of xgboost". This PR removes it as an option so the code doesn't just crash.

[github-actions]

Before [051369d]	After [1985267]	Ratio	Benchmark (Parameter)
failed	failed	n/a	benchmarks.peakmem_learn_loop('KNN')
failed	failed	n/a	benchmarks.peakmem_learn_loop('RandomForest')
170±2ms	175±4ms	1.03	benchmarks.time_learn_loop('KNN', 'RandomSampling')
174±2ms	179±3ms	1.02	benchmarks.time_learn_loop('KNN', 'UncSampling')
2.67±0.02s	2.68±0s	1.00	benchmarks.time_learn_loop('RandomForest', 'RandomSampling')
139±1ms	136±2ms	0.98	benchmarks.time_feature_creation
2.71±0.02s	2.66±0.01s	0.98	benchmarks.time_learn_loop('RandomForest', 'UncSampling')

Click here to view all benchmarks.

[drewoldag]

Looks good. Thank you.