Added Bayes backend benchmarks

[ibnesayeed]

Comparing the performance of training, untraining, and classification tasks of Memory and Redis backends. Followup from #95.

The following benchmarks was generated on 1, 10, 100, 1000, and 5000 records.

$  rake bench NOPROGRESS=T
/usr/local/bin/ruby -w -I"lib:lib" -I"/usr/local/bundle/gems/rake-12.0.0/lib" "/usr/local/bundle/gems/rake-12.0.0/lib/rake/rake_test_loader.rb" "test/bayes/bayesian_memory_benchmark.rb" "test/bayes/bayesian_redis_benchmark.rb" 
Run options: --seed 20883

# Running:

redis_bench_train	 0.003239	 0.027831	 0.154580	 1.444252	 6.938907
.redis_bench_train_untrain	 0.006432	 0.062580	 0.498467	 4.458813	21.229394
.redis_bench_train_classify	 0.004220	 0.038421	 0.322356	 3.319935	17.069817
.memory_bench_train	 0.000167	 0.000885	 0.004427	 0.037115	 0.180178
.memory_bench_train_untrain	 0.000185	 0.001424	 0.012125	 0.078261	 0.392870
.memory_bench_train_classify	 0.000220	 0.001524	 0.009205	 0.086727	 0.436754
.

Finished in 57.102266s, 0.1051 runs/s, 0.1051 assertions/s.

6 runs, 6 assertions, 0 failures, 0 errors, 0 skips

[ibnesayeed]

Here are some plots based on the above data. Generated using this spreadsheet.