diff --git a/Ordinal/experiments_lr_vs_ordlr.py b/Ordinal/experiments_lr_vs_ordlr.py
index ff4f56a..a6712ca 100644
--- a/Ordinal/experiments_lr_vs_ordlr.py
+++ b/Ordinal/experiments_lr_vs_ordlr.py
@@ -3,11 +3,11 @@ import quapy as qp
 import os
 from sklearn.linear_model import LogisticRegression
 from sklearn.preprocessing import StandardScaler
-from Ordinal.model import RegressionQuantification, LogisticAT, LogisticSE, LogisticIT, LAD, OrdinalRidge
+from Ordinal.model import LogisticAT, LogisticSE, LogisticIT, LAD, OrdinalRidge #, RegressionQuantification
 from quapy.method.aggregative import PACC, CC, EMQ, PCC, ACC
 from os.path import join
 from utils import load_samples_folder, load_single_sample_pkl
-from evaluation import nmd, mnmd
+from Ordinal.evaluation import nmd, mnmd
 from tqdm import tqdm
 
 
@@ -37,11 +37,11 @@ def quantifiers():
     yield 'PACC(OLR-AT)', PACC(LogisticAT()), params_OLR
     yield 'SLD(OLR-AT)', EMQ(LogisticAT()), params_OLR
 
-    yield 'CC(OLR-SE)', CC(LogisticSE()), params_OLR
-    yield 'PCC(OLR-SE)', PCC(LogisticSE()), params_OLR
-    yield 'ACC(OLR-SE)', ACC(LogisticSE()), params_OLR
-    yield 'PACC(OLR-SE)', PACC(LogisticSE()), params_OLR
-    yield 'SLD(OLR-SE)', EMQ(LogisticSE()), params_OLR
+    # yield 'CC(OLR-SE)', CC(LogisticSE()), params_OLR
+    # yield 'PCC(OLR-SE)', PCC(LogisticSE()), params_OLR
+    # yield 'ACC(OLR-SE)', ACC(LogisticSE()), params_OLR
+    # yield 'PACC(OLR-SE)', PACC(LogisticSE()), params_OLR
+    # yield 'SLD(OLR-SE)', EMQ(LogisticSE()), params_OLR
 
     yield 'CC(OLR-IT)', CC(LogisticIT()), params_OLR
     yield 'PCC(OLR-IT)', PCC(LogisticIT()), params_OLR
@@ -53,6 +53,7 @@ def quantifiers():
     # regression-based ordinal regression (see https://pythonhosted.org/mord/) 
     yield 'CC(LAD)', CC(LAD()), params_SVR
     yield 'ACC(LAD)', ACC(LAD()), params_SVR
+
     yield 'CC(ORidge)', CC(OrdinalRidge()), params_Ridge
     yield 'ACC(ORidge)', ACC(OrdinalRidge()), params_Ridge
 
@@ -60,7 +61,7 @@ def quantifiers():
 def run_experiment(params):
     qname, q, param_grid = params
     qname += posfix
-    resultfile = join(resultpath, f'{qname}.all.csv')
+    resultfile = join(resultpath, f'{qname}.all.APP-OQ.csv')
     if os.path.exists(resultfile):
         print(f'result file {resultfile} already exists: continue')
         return None
@@ -105,16 +106,16 @@ def run_experiment(params):
     print(f'{qname}: {mean_nmd:.4f} +-{std_nmd:.4f}')
     report.to_csv(resultfile, index=False)
 
-    print('[learning regressor-based adjustment]')
-    q = RegressionQuantification(q.best_model(), val_samples_generator=load_dev_samples)
-    q.fit(None)
+    # print('[learning regressor-based adjustment]')
+    # q = RegressionQuantification(q.best_model(), val_samples_generator=load_dev_samples)
+    # q.fit(None)
 
-    report = qp.evaluation.gen_prevalence_report(q, gen_fn=load_test_samples, error_metrics=[nmd])
-    mean_nmd = report['nmd'].mean()
-    std_nmd = report['nmd'].std()
-    print(f'[{qname} regression-correction] {mean_nmd:.4f} +-{std_nmd:.4f}')
-    resultfile = join(resultpath, f'{qname}.all.reg.csv')
-    report.to_csv(resultfile, index=False)
+    # report = qp.evaluation.gen_prevalence_report(q, gen_fn=load_test_samples, error_metrics=[nmd])
+    # mean_nmd = report['nmd'].mean()
+    # std_nmd = report['nmd'].std()
+    # print(f'[{qname} regression-correction] {mean_nmd:.4f} +-{std_nmd:.4f}')
+    # resultfile = join(resultpath, f'{qname}.all.reg.csv')
+    # report.to_csv(resultfile, index=False)
 
     return hyperparams
 
diff --git a/Ordinal/utils.py b/Ordinal/utils.py
index 182851d..fc74962 100644
--- a/Ordinal/utils.py
+++ b/Ordinal/utils.py
@@ -12,6 +12,9 @@ import quapy as qp
 from quapy.data import LabelledCollection
 
 
+def jaggedness(p):
+    return (1/min(6, len(p)+1)) * sum((-p_prev + 2*p_i - p_next)**2 for p_prev, p_i, p_next in zip(p[:-2], p[1:-1], p[2:]))
+
 
 def load_simple_sample_npytxt(parentdir, filename, classes=None):
     samplepath = join(parentdir, filename+'.txt')