KnowledgeCaptureAndDiscovery · dgarijo · Feb 26, 2026 · Dec 10, 2025 · Dec 11, 2025 · Dec 11, 2025
diff --git a/experiments/create_models.py b/experiments/create_models.py
@@ -36,13 +36,13 @@
     'cvbb': make_pipeline(CountVectorizer(), BernoulliNB()),
     'cvlr': make_pipeline(CountVectorizer(), LogisticRegression(solver='liblinear')),
     'cvnb': make_pipeline(CountVectorizer(), MultinomialNB()),
-    'tfada': make_pipeline(TfidfVectorizer(), AdaBoostClassifier()),  #(max_depth=3, random_state=0))
+    'tfada': make_pipeline(TfidfVectorizer(), AdaBoostClassifier(algorithm="SAMME")),  #(max_depth=3, random_state=0))
     'tfdtc': make_pipeline(CountVectorizer(), DecisionTreeClassifier()),
     'tflr': make_pipeline(TfidfVectorizer(), LogisticRegression(solver='liblinear')),
     'tfnb': make_pipeline(TfidfVectorizer(), MultinomialNB()),
     'tfrfc': make_pipeline(TfidfVectorizer(), RandomForestClassifier()),  #(max_depth=3, random_state=0))
     'tfsgd': make_pipeline(TfidfVectorizer(), SGDClassifier(loss='log_loss')),
-    'tfxgb': make_pipeline(TfidfVectorizer(), XGBClassifier(use_label_encoder=False,eval_metric="logloss"))
+    'tfxgb': make_pipeline(TfidfVectorizer(), XGBClassifier(eval_metric="logloss"))
 }
 #'tfper': make_pipeline(TfidfVectorizer(), Perceptron(tol=1e-3, random_state=0)),
 evaluation_names = ('cvlr', 'tflr', 'tfnb', 'cvnb', 'cvbb', 'tfsgd', 'tfxgb', 'tfper', 'tfrfc', 'tfdtc', 'tfada')

diff --git a/experiments/ranking/citation_classifier.csv b/experiments/ranking/citation_classifier.csv
@@ -1,12 +1,11 @@
 sklearn Primitive - citation,Hyperparameters,Input Data Used,Accuracy,Precision,Recall,F-measure,Pickle ID
-"sklearnpipeline(CountVectorizer, BernoulliBayes)", -,Allen,0.754,0.675,0.993,0.803,citcvbb.p
-"sklearnpipeline(CountVectorizer, LogisticRegression)", - ,Allen,0.922,0.912,0.938,0.924,citcvlr.p
-"sklearnpipeline(CountVectorizer, NaiveBayes)", -,Allen,0.927,0.909,0.952,0.929,citcvnb.p
-"sklearnpipeline(TFIDFVectorizer, AdaBoostClassifier)", -,Allen,0.886,0.948,0.821,0.877,cittfada.p
-"sklearnpipeline(TFIDFVectorizer, DecisionTreeClassifier)", -,Allen,0.879,0.95,0.8,0.868,cittfdtc.p
-"sklearnpipeline(TFIDFVectorizer, LogisticRegression)", - ,Allen,0.938,0.958,0.917,0.936,cittflr.p
-"sklearnpipeline(TFIDFVectorizer, NaiveBayes)", - ,Allen,0.92,0.91,0.934,0.922,cittfnb.p
-"sklearnpipeline(TFIDFVectorizer, Perceptron)", -,Allen,0.907,0.899,0.921,0.909,cittfper.p
-"sklearnpipeline(TFIDFVectorizer, RandomForestClassifier)", -,Allen,0.879,0.864,0.907,0.882,cittfrfc.p
-"sklearnpipeline(TFIDFVectorizer, StochasticGradientDescent)",loss = 'log',Allen,0.939,0.957,0.921,0.938,cittfsgd.p
-"sklearnpipeline(TFIDFVectorizer, XGBClassifier)", -,Allen,0.839,0.906,0.759,0.825,cittfxgb.p
+"sklearnpipeline(CountVectorizer, BernoulliBayes)", -,Allen,0.765,0.685,0.983,0.807,citcvbb.p
+"sklearnpipeline(CountVectorizer, LogisticRegression)", - ,Allen,0.926,0.908,0.948,0.927,citcvlr.p
+"sklearnpipeline(CountVectorizer, NaiveBayes)", -,Allen,0.945,0.932,0.962,0.946,citcvnb.p
+"sklearnpipeline(TFIDFVectorizer, AdaBoostClassifier)", -,Allen,0.768,0.818,0.755,0.757,cittfada.p
+"sklearnpipeline(TFIDFVectorizer, DecisionTreeClassifier)", -,Allen,0.846,0.877,0.821,0.843,cittfdtc.p
+"sklearnpipeline(TFIDFVectorizer, LogisticRegression)", - ,Allen,0.933,0.964,0.9,0.93,cittflr.p
+"sklearnpipeline(TFIDFVectorizer, NaiveBayes)", - ,Allen,0.936,0.921,0.959,0.938,cittfnb.p
+"sklearnpipeline(TFIDFVectorizer, RandomForestClassifier)", -,Allen,0.879,0.866,0.9,0.882,cittfrfc.p
+"sklearnpipeline(TFIDFVectorizer, StochasticGradientDescent)",loss = 'log',Allen,0.945,0.971,0.917,0.943,cittfsgd.p
+"sklearnpipeline(TFIDFVectorizer, XGBClassifier)", -,Allen,0.832,0.847,0.831,0.831,cittfxgb.p
diff --git a/experiments/ranking/description_classifier.csv b/experiments/ranking/description_classifier.csv
@@ -1,12 +1,11 @@
 sklearn Primitive - description,Hyperparameters,Input Data Used,Accuracy,Precision,Recall,F-measure,Pickle ID
-"sklearnpipeline(CountVectorizer, BernoulliBayes)", -,Allen,0.733,0.914,0.516,0.657,descvbb.p
-"sklearnpipeline(CountVectorizer, LogisticRegression)", - ,Allen,0.813,0.854,0.758,0.801,descvlr.p
-"sklearnpipeline(CountVectorizer, NaiveBayes)", -,Allen,0.784,0.716,0.954,0.816,descvnb.p
-"sklearnpipeline(TFIDFVectorizer, AdaBoostClassifier)", -,Allen,0.731,0.73,0.737,0.733,destfada.p
-"sklearnpipeline(TFIDFVectorizer, DecisionTreeClassifier)", -,Allen,0.686,0.701,0.648,0.67,destfdtc.p
-"sklearnpipeline(TFIDFVectorizer, LogisticRegression)", - ,Allen,0.802,0.788,0.836,0.809,destflr.p
-"sklearnpipeline(TFIDFVectorizer, NaiveBayes)", - ,Allen,0.775,0.695,0.982,0.814,destfnb.p
-"sklearnpipeline(TFIDFVectorizer, Perceptron)", -,Allen,0.804,0.777,0.854,0.814,destfper.p
-"sklearnpipeline(TFIDFVectorizer, RandomForestClassifier)", -,Allen,0.745,0.744,0.751,0.747,destfrfc.p
-"sklearnpipeline(TFIDFVectorizer, StochasticGradientDescent)",loss = 'log',Allen,0.815,0.799,0.843,0.82,destfsgd.p
-"sklearnpipeline(TFIDFVectorizer, XGBClassifier)", -,Allen,0.72,0.729,0.726,0.724,destfxgb.p
+"sklearnpipeline(CountVectorizer, BernoulliBayes)", -,Allen,0.764,0.965,0.551,0.699,descvbb.p
+"sklearnpipeline(CountVectorizer, LogisticRegression)", - ,Allen,0.822,0.857,0.778,0.814,descvlr.p
+"sklearnpipeline(CountVectorizer, NaiveBayes)", -,Allen,0.789,0.712,0.975,0.823,descvnb.p
+"sklearnpipeline(TFIDFVectorizer, AdaBoostClassifier)", -,Allen,0.768,0.778,0.757,0.767,destfada.p
+"sklearnpipeline(TFIDFVectorizer, DecisionTreeClassifier)", -,Allen,0.739,0.724,0.782,0.75,destfdtc.p
+"sklearnpipeline(TFIDFVectorizer, LogisticRegression)", - ,Allen,0.822,0.789,0.885,0.833,destflr.p
+"sklearnpipeline(TFIDFVectorizer, NaiveBayes)", - ,Allen,0.768,0.69,0.984,0.81,destfnb.p
+"sklearnpipeline(TFIDFVectorizer, RandomForestClassifier)", -,Allen,0.799,0.812,0.786,0.798,destfrfc.p
+"sklearnpipeline(TFIDFVectorizer, StochasticGradientDescent)",loss = 'log',Allen,0.834,0.807,0.885,0.843,destfsgd.p
+"sklearnpipeline(TFIDFVectorizer, XGBClassifier)", -,Allen,0.768,0.75,0.815,0.78,destfxgb.p
diff --git a/experiments/ranking/installation_classifier.csv b/experiments/ranking/installation_classifier.csv
@@ -1,12 +1,11 @@
 sklearn Primitive - installation,Hyperparameters,Input Data Used,Accuracy,Precision,Recall,F-measure,Pickle ID
-"sklearnpipeline(CountVectorizer, BernoulliBayes)", -,Allen,0.76,0.678,0.991,0.805,inscvbb.p
-"sklearnpipeline(CountVectorizer, LogisticRegression)", - ,Allen,0.894,0.874,0.923,0.897,inscvlr.p
-"sklearnpipeline(CountVectorizer, NaiveBayes)", -,Allen,0.898,0.853,0.961,0.904,inscvnb.p
-"sklearnpipeline(TFIDFVectorizer, AdaBoostClassifier)", -,Allen,0.797,0.882,0.685,0.771,instfada.p
-"sklearnpipeline(TFIDFVectorizer, DecisionTreeClassifier)", -,Allen,0.825,0.841,0.807,0.822,instfdtc.p
-"sklearnpipeline(TFIDFVectorizer, LogisticRegression)", - ,Allen,0.894,0.938,0.845,0.889,instflr.p
-"sklearnpipeline(TFIDFVectorizer, NaiveBayes)", - ,Allen,0.897,0.851,0.965,0.904,instfnb.p
-"sklearnpipeline(TFIDFVectorizer, Perceptron)", -,Allen,0.887,0.887,0.887,0.887,instfper.p
-"sklearnpipeline(TFIDFVectorizer, RandomForestClassifier)", -,Allen,0.87,0.854,0.894,0.873,instfrfc.p
-"sklearnpipeline(TFIDFVectorizer, StochasticGradientDescent)",loss = 'log',Allen,0.903,0.922,0.882,0.901,instfsgd.p
-"sklearnpipeline(TFIDFVectorizer, XGBClassifier)", -,Allen,0.833,0.874,0.781,0.824,instfxgb.p
+"sklearnpipeline(CountVectorizer, BernoulliBayes)", -,Allen,0.768,0.688,0.985,0.81,inscvbb.p
+"sklearnpipeline(CountVectorizer, LogisticRegression)", - ,Allen,0.903,0.88,0.936,0.906,inscvlr.p
+"sklearnpipeline(CountVectorizer, NaiveBayes)", -,Allen,0.91,0.872,0.962,0.915,inscvnb.p
+"sklearnpipeline(TFIDFVectorizer, AdaBoostClassifier)", -,Allen,0.713,0.824,0.63,0.672,instfada.p
+"sklearnpipeline(TFIDFVectorizer, DecisionTreeClassifier)", -,Allen,0.846,0.883,0.798,0.838,instfdtc.p
+"sklearnpipeline(TFIDFVectorizer, LogisticRegression)", - ,Allen,0.891,0.924,0.852,0.886,instflr.p
+"sklearnpipeline(TFIDFVectorizer, NaiveBayes)", - ,Allen,0.909,0.867,0.968,0.914,instfnb.p
+"sklearnpipeline(TFIDFVectorizer, RandomForestClassifier)", -,Allen,0.873,0.856,0.897,0.876,instfrfc.p
+"sklearnpipeline(TFIDFVectorizer, StochasticGradientDescent)",loss = 'log',Allen,0.914,0.932,0.893,0.912,instfsgd.p
+"sklearnpipeline(TFIDFVectorizer, XGBClassifier)", -,Allen,0.839,0.88,0.786,0.83,instfxgb.p
diff --git a/experiments/ranking/invocation_classifier.csv b/experiments/ranking/invocation_classifier.csv
@@ -1,12 +1,11 @@
 sklearn Primitive - invocation,Hyperparameters,Input Data Used,Accuracy,Precision,Recall,F-measure,Pickle ID
-"sklearnpipeline(CountVectorizer, BernoulliBayes)", -,Allen,0.755,0.681,0.962,0.797,invcvbb.p
-"sklearnpipeline(CountVectorizer, LogisticRegression)", - ,Allen,0.858,0.828,0.907,0.865,invcvlr.p
-"sklearnpipeline(CountVectorizer, NaiveBayes)", -,Allen,0.885,0.874,0.899,0.886,invcvnb.p
-"sklearnpipeline(TFIDFVectorizer, AdaBoostClassifier)", -,Allen,0.771,0.74,0.836,0.785,invtfada.p
-"sklearnpipeline(TFIDFVectorizer, DecisionTreeClassifier)", -,Allen,0.766,0.757,0.784,0.77,invtfdtc.p
-"sklearnpipeline(TFIDFVectorizer, LogisticRegression)", - ,Allen,0.869,0.87,0.868,0.869,invtflr.p
-"sklearnpipeline(TFIDFVectorizer, NaiveBayes)", - ,Allen,0.89,0.878,0.907,0.892,invtfnb.p
-"sklearnpipeline(TFIDFVectorizer, Perceptron)", -,Allen,0.848,0.845,0.853,0.849,invtfper.p
-"sklearnpipeline(TFIDFVectorizer, RandomForestClassifier)", -,Allen,0.823,0.807,0.85,0.827,invtfrfc.p
-"sklearnpipeline(TFIDFVectorizer, StochasticGradientDescent)",loss = 'log',Allen,0.879,0.878,0.881,0.879,invtfsgd.p
-"sklearnpipeline(TFIDFVectorizer, XGBClassifier)", -,Allen,0.803,0.771,0.863,0.815,invtfxgb.p
+"sklearnpipeline(CountVectorizer, BernoulliBayes)", -,Allen,0.758,0.684,0.965,0.8,invcvbb.p
+"sklearnpipeline(CountVectorizer, LogisticRegression)", - ,Allen,0.86,0.833,0.903,0.866,invcvlr.p
+"sklearnpipeline(CountVectorizer, NaiveBayes)", -,Allen,0.892,0.879,0.91,0.894,invcvnb.p
+"sklearnpipeline(TFIDFVectorizer, AdaBoostClassifier)", -,Allen,0.731,0.701,0.805,0.749,invtfada.p
+"sklearnpipeline(TFIDFVectorizer, DecisionTreeClassifier)", -,Allen,0.782,0.778,0.79,0.784,invtfdtc.p
+"sklearnpipeline(TFIDFVectorizer, LogisticRegression)", - ,Allen,0.875,0.881,0.868,0.874,invtflr.p
+"sklearnpipeline(TFIDFVectorizer, NaiveBayes)", - ,Allen,0.895,0.887,0.905,0.896,invtfnb.p
+"sklearnpipeline(TFIDFVectorizer, RandomForestClassifier)", -,Allen,0.826,0.811,0.853,0.831,invtfrfc.p
+"sklearnpipeline(TFIDFVectorizer, StochasticGradientDescent)",loss = 'log',Allen,0.88,0.886,0.874,0.88,invtfsgd.p
+"sklearnpipeline(TFIDFVectorizer, XGBClassifier)", -,Allen,0.799,0.771,0.853,0.81,invtfxgb.p
diff --git a/experiments/trained_models/citcvbb.p b/experiments/trained_models/citcvbb.p
diff --git a/experiments/trained_models/citcvlr.p b/experiments/trained_models/citcvlr.p
diff --git a/experiments/trained_models/citcvnb.p b/experiments/trained_models/citcvnb.p
diff --git a/experiments/trained_models/cittfada.p b/experiments/trained_models/cittfada.p
diff --git a/experiments/trained_models/cittfdtc.p b/experiments/trained_models/cittfdtc.p
diff --git a/experiments/trained_models/cittflr.p b/experiments/trained_models/cittflr.p
diff --git a/experiments/trained_models/cittfnb.p b/experiments/trained_models/cittfnb.p
diff --git a/experiments/trained_models/cittfrfc.p b/experiments/trained_models/cittfrfc.p
diff --git a/experiments/trained_models/cittfsgd.p b/experiments/trained_models/cittfsgd.p
diff --git a/experiments/trained_models/cittfxgb.p b/experiments/trained_models/cittfxgb.p
diff --git a/experiments/trained_models/descvbb.p b/experiments/trained_models/descvbb.p
diff --git a/experiments/trained_models/descvlr.p b/experiments/trained_models/descvlr.p
diff --git a/experiments/trained_models/descvnb.p b/experiments/trained_models/descvnb.p
diff --git a/experiments/trained_models/destfada.p b/experiments/trained_models/destfada.p
diff --git a/experiments/trained_models/destfdtc.p b/experiments/trained_models/destfdtc.p
diff --git a/experiments/trained_models/destflr.p b/experiments/trained_models/destflr.p
diff --git a/experiments/trained_models/destfnb.p b/experiments/trained_models/destfnb.p
diff --git a/experiments/trained_models/destfrfc.p b/experiments/trained_models/destfrfc.p
diff --git a/experiments/trained_models/destfsgd.p b/experiments/trained_models/destfsgd.p
diff --git a/experiments/trained_models/destfxgb.p b/experiments/trained_models/destfxgb.p
diff --git a/experiments/trained_models/inscvbb.p b/experiments/trained_models/inscvbb.p
diff --git a/experiments/trained_models/inscvlr.p b/experiments/trained_models/inscvlr.p
diff --git a/experiments/trained_models/inscvnb.p b/experiments/trained_models/inscvnb.p
diff --git a/experiments/trained_models/instfada.p b/experiments/trained_models/instfada.p
diff --git a/experiments/trained_models/instfdtc.p b/experiments/trained_models/instfdtc.p
diff --git a/experiments/trained_models/instflr.p b/experiments/trained_models/instflr.p
diff --git a/experiments/trained_models/instfnb.p b/experiments/trained_models/instfnb.p
diff --git a/experiments/trained_models/instfrfc.p b/experiments/trained_models/instfrfc.p
diff --git a/experiments/trained_models/instfsgd.p b/experiments/trained_models/instfsgd.p
diff --git a/experiments/trained_models/instfxgb.p b/experiments/trained_models/instfxgb.p
diff --git a/experiments/trained_models/invcvbb.p b/experiments/trained_models/invcvbb.p
diff --git a/experiments/trained_models/invcvlr.p b/experiments/trained_models/invcvlr.p
diff --git a/experiments/trained_models/invcvnb.p b/experiments/trained_models/invcvnb.p
diff --git a/experiments/trained_models/invtfada.p b/experiments/trained_models/invtfada.p
diff --git a/experiments/trained_models/invtfdtc.p b/experiments/trained_models/invtfdtc.p
diff --git a/experiments/trained_models/invtflr.p b/experiments/trained_models/invtflr.p
diff --git a/experiments/trained_models/invtfnb.p b/experiments/trained_models/invtfnb.p
diff --git a/experiments/trained_models/invtfrfc.p b/experiments/trained_models/invtfrfc.p
diff --git a/experiments/trained_models/invtfsgd.p b/experiments/trained_models/invtfsgd.p
diff --git a/experiments/trained_models/invtfxgb.p b/experiments/trained_models/invtfxgb.p