Simplifying decision trees

1987

Citations in the corpus

A Survey of Methods for Explaining Black Box Models (2018, background): “Indeed, a decision tree can be linearized into a set of decision rules with the if-then form [77,78,26]:”
Conservative Q-Improvement: Reinforcement Learning for an Interpretable Decision-Tree Policy (2019, background): “In an AI context speciﬁcally, it has been suggested that huge decision trees are insufﬁciently comprehensible to humans, and that simplifying trees is important for humans to interact with the representation [Quinlan, 1987].”
Methods and Models for Interpretable Linear Classification (2014, background): “The interpretability of transparent models is usually improved by tuning sparsity [107, 120, 29, 50, 89, 19, 59, 47], by ensuring monotonic relationships between certain input variables and the predicted outcome [10, 87, 114, 77], and by restricting coefficients to a small set of integer values [23,”
Multivariate Decision Trees (1995, methodology): “A full tree is grown, which classiies all the training instances correctly and then subtrees are pruned back to reduce future classiication errors (Breiman, Friedman, Olshen & Stone, 1984; Quinlan, 1987).”
Induction of ripple-down rules applied to modeling large databases (1995, background): “There are some minor internal refinements to cope with known problems of empirical induction as described by Quinlan (1987) and Cendrowska (1987).”
STATLOG: COMPARISON OF CLASSIFICATION ALGORITHMS ON LARGE REAL-WORLD PROBLEMS (1995, unknown): “They can be divided into four categories: Symbolic classiiers, including C4.5 Quinlan, 1987a], NewID, AC 2 , CN2 Clark and Boswell, 1991], ITrule Goodman and Smyth, 1989], CART Breiman et al., 1984], and INDCART; Statistical classiiers, including k-nearest-neighbor, \naive" Bayes, linear discriminan”
DATA MINING TECHNIQUES (2003, background): “More details on tree pruning can be found in [55, 56, 65, 67]; we will focus on the build tree step.”
Learning Decision Trees Using the Area Under the ROC Curve (2002, methodology): “The postpruning method used is the “Pessimistic Error Pruning” introduced by (Quinlan 1987).”
Extracting refined rules from knowledge-based neural networks (2004, unknown): “Both Saito and Nakano (1988) and Fu (1991) report a method similar to our SUBSET algorithm.”
A Survey of Methods for Scaling Up Inductive Algorithms (1999, methodology): “The most common technique for producing high-accuracy rule sets, known as reduced-error pruning , is to grow rules via one algorithm or another, and then prune the rules in order to increase accuracy (Quinlan, 1987).”
Rule Induction with CN2: Some Recent Improvements (1991, unknown): “Also, comparisons with Quinlan's C4.5 Quinlan et al., 1987, Quinlan, 1987] were performed.”
Evaluation of random forest and adaboost tree-based ensemble classification and spectral band selection for ecotope mapping using airborne hyperspectral imagery (2008, methodology): “Some common pruning methods are discussed in Quinlan (1987) and Mingers (1989b).”
A review of associative classification mining (2007, unknown): “The ADT uses pessimistic error pruning (Quinlan, 1987), which constructs a decision-tree-like structure, known as an ADT-tree, using the generated CARs and places general rules at the higher levels and specific rules at the lower levels of the tree.”
A System for Induction of Oblique Decision Trees (1994, background): “Many studies have found that judicious pruning results in both smaller and more accurate classi ers, for decision trees as well as other types of machine learning systems (Quinlan, 1987; Niblett, 1986; Cestnik, Kononenko, & Bratko, 1987; Kodrato & Manago, 1987; Cohen, 1993; Hassibi & Stork, 1993; Wo”
Top-down induction of decision trees classifiers - a survey (2005, unknown): “A LGORITHMIC F RAMEWORK FOR D ECISION T REES Decision Tree inducers are algorithms that automatically construct a decision tree from a given dataset.”
Building Decision Trees with Constraints (2003, methodology): “The second class of pruning algorithms, pessimistic pruning (Quinlan, 1987), do not require separate pruning data, and are computationally inexpensive.”