This tutorial shows how to apply General Regressor for a regression problem of e2006 dataset.

Training

set mapred.reduce.tasks=32;

drop table e2006tfidf_generic_model;
create table e2006tfidf_generic_model as
select 
 feature,
 avg(weight) as weight
from 
 (select 
     train_regressor(
       add_bias(features), target,
       '-loss squaredloss -opt AdamHD -reg No -iters 20'
     ) as (feature, weight)
  from 
     e2006tfidf_train_x3
 ) t 
group by feature;

-- reset to the default setting
set mapred.reduce.tasks=-1;

prediction

create or replace view e2006tfidf_generic_predict
as
select
  t.rowid, 
  sum(m.weight * t.value) as predicted
from 
  e2006tfidf_test_exploded t LEFT OUTER JOIN
  e2006tfidf_generic_model m ON (t.feature = m.feature)
group by
  t.rowid;

evaluation

WITH submit as (
  select 
    t.target as actual, 
    p.predicted as predicted
  from 
    e2006tfidf_test t
    JOIN e2006tfidf_generic_predict p 
      on (t.rowid = p.rowid)
)
select 
   rmse(predicted, actual) as RMSE,
   mse(predicted, actual) as MSE, 
   mae(predicted, actual) as MAE,
   r2(predicted, actual) as R2
from 
   submit;