[MADLIB-641] SVM Classification: MADlib has worse score than libsvm on data set splice

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Abandoned
• Fix Version/s: v1.9
• Labels: severity_set
• Assignee: Rahul Iyer
• Reporter: Jiali Yao
• Created: 06/Sep/12 08:54
• Updated: 22/Mar/18 18:46
• Resolved: 22/Mar/18 18:46

Description

kernel    Data Sets	MADlib(Para = true)  MADlib(Para = true)   Libsvm	Max(Madlib)/libsvm
linea      splice	0.526	             0.556	           0.848	0.656
gaussina  splice	0.520	             0.480              	0.904	0.575

Linear case

-- method: svm_cls_linear_ds_0_6_lsvm_classification_0
SELECT madlib.lsvm_classification
                        ( 'madlibtestdata.svm_splice'::text     --input_table
                        , 'madlibtestresult.cls_model_table'::text    --model_table
                        , 'true'::boolean       --parallel
                        , 'false'::boolean        --verbose
                        , '0.1'::float8            --eta
                        , '0.001'::float8            --reg

                   ) AS q;
-- method: svm_cls_linear_ds_0_6_svm_cls_predict_batch_score_2
SELECT madlibtestdata.svm_cls_predict_batch_score
                        ( 'madlibtestresult.cls_model_table'::text     --model_table
                        , 'madlibtestresult.cls_output_table'::text  --output_table
                        , 'madlibtestdata.svm_splice_test'::text     --input_table
                        , 'true'::boolean        --parallel
                        , 'true'::boolean        --linear
                   ) AS score;

-- method: svm_cls_linear_ds_0_6_svm_drop_model_3
SELECT madlib.svm_drop_model
                        ('madlibtestresult.cls_model_table'::text     --model_table
                   ) AS q;

Gaussian case

-- method: svm_cls_gaussian_ds_0_6_svm_classification_0
SELECT madlib.svm_classification
                        ( 'madlibtestdata.svm_splice'::text     --input_table
                        , 'madlibtestresult.cls_model_table'::text    --model_table
                        , 'true'::boolean       --parallel
                        , 'madlibtestdata.svm_gaussian_splice'::text    --kernel_func
                        , 'false'::boolean        --verbose
                        , '0.01'::float8            --eta
                        , '0.005'::float8             --nu
                   ) AS q;


-- method: svm_cls_gaussian_ds_0_6_svm_cls_predict_batch_score_2
SELECT madlibtestdata.svm_cls_predict_batch_score
                        ( 'madlibtestresult.cls_model_table'::text     --model_table
                        , 'madlibtestresult.cls_output_table'::text  --output_table
                        , 'madlibtestdata.svm_splice_test'::text     --input_table
                        , 'false'::boolean        --parallel
                        , 'false'::boolean        --linear
                   ) AS score;