Estimating Quantities: Comparing Simple Heuristics and Machine Learning Algorithms

Woike, Jan K.; Hoffrage, Ulrich; Hertwig, Ralph

doi:10.1007/978-3-642-33266-1_60

Cited by 3 publications

(3 citation statements)

References 11 publications

(8 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Previous research for other tasks, such as paired comparison (Czerlinski, Gigerenzer, & Goldstein, 1999; Martignon & Hoffrage, 2002) or estimation (Hertwig, Hoffrage, & Sparr, 2012; Woike, Hoffrage, & Hertwig, 2012), has shown that simple heuristics, when making inferences out of sample, can outperform models that are optimal when it comes to making inferences within the same sample for which the strategies fitted their parameters. We found the same pattern in our simulations: When making classifications out of sample with multiple cues, the two heuristic approaches, naïve Bayes and some of the fast-and-frugal trees, outperformed the model that was normative for the case of fitting known data, namely classification based on natural frequencies (or, equivalently, profile memorization).…”

Section: Discussionmentioning

confidence: 99%

Integrating and testing natural frequencies, naïve Bayes, and fast-and-frugal trees.

Woike¹,

Hoffrage²,

Martignon³

2017

Decision

Self Cite

View full text Add to dashboard Cite

This article relates natural frequency representations of cue-criterion relationships to fast-and-frugal heuristics for inferences based on multiple cues. In the conceptual part of this work, three approaches to classification are compared to one another: The first uses a natural Bayesian classification scheme, based on profile memorization and natural frequencies. The second is based on Naïve Bayes, a heuristic that assumes conditional independence between cues (given the criterion). The third approach is to construct fast-and-frugal classification trees, which can be conceived as pruned versions of diagnostic natural frequency trees. Fastand-frugal trees can be described as lexicographic classifiers but can also be related to another fundamental class of models, namely linear models. Linear classifiers with fixed thresholds and noncompensatory weights coincide with fast-and-frugal trees-not as processes but in their output. Various heuristic principles for tree construction are proposed. In the second, empirical part of this article, the classification performance of the three approaches when making inferences under uncertainty (i.e., out of sample) is evaluated in 11 medical data sets in terms of Receiver Operating Characteristics (ROC) diagrams and predictive accuracy. Results show that the two heuristic approaches, Naïve Bayes and fast-and-frugal trees, generally outperformed the model that is normative when fitting known data, namely classification based on natural frequencies (or, equivalently, profile memorization). The success of fast-and-frugal trees is grounded in their ecological rationality: their construction principles can exploit the structure of information in the data sets. Finally, implications, applications, limitations, and possible extensions of this work are discussed.

show abstract

Section: Discussionmentioning

confidence: 99%

Integrating and testing natural frequencies, naïve Bayes, and fast-and-frugal trees.

Woike¹,

Hoffrage²,

Martignon³

2017

Decision

Self Cite

View full text Add to dashboard Cite

show abstract

“…So how well do FFTs predict data relative to more complex algorithms when there are no computational restrictions? Prior research has suggested that FFTs can predict data as well as algorithms such as logistic regression and standard decision trees (Martignon et al, 2008;Woike, Hoffrage & Hertwig, 2012;Woike et al, 2017). However, logistic regression (in a non-regularised form) and standard decision 11For this example, 100 FFTs were constructed from 100 simulations using 50% of the full heart disease data.…”

Section: Prediction Simulationmentioning

confidence: 99%

FFTrees: A toolbox to create, visualize, and evaluate fast-and-frugal decision trees

Phillips¹,

Neth

Woike

et al. 2017

Judgm. decis. mak.

View full text Add to dashboard Cite

Fast-and-frugal trees (FFTs) are simple algorithms that facilitate efficient and accurate decisions based on limited information. But despite their successful use in many applied domains, there is no widely available toolbox that allows anyone to easily create, visualize, and evaluate FFTs. We fill this gap by introducing the R package FFTrees. In this paper, we explain how FFTs work, introduce a new class of algorithms called fan for constructing FFTs, and provide a tutorial for using the FFTrees package. We then conduct a simulation across ten real-world datasets to test how well FFTs created by FFTrees can predict data. Simulation results show that FFTs created by FFTrees can predict data as well as popular classification algorithms such as regression and random forests, while remaining simple enough for anyone to understand and use.

show abstract

“…Heuristic categorization and the visualization of fast-and-frugal trees was performed according to established heuristic tree construction algorithms as previously described [ 36 , 37 ].…”

Section: Methodsmentioning

confidence: 99%

Complex Oncological Decision-Making Utilizing Fast-and-Frugal Trees in a Community Setting—Role of Academic and Hybrid Modeling

Salgia

Mambetsariev

Tan

et al. 2020

JCM

View full text Add to dashboard Cite

Non-small cell lung cancer is a devastating disease and with the advent of targeted therapies and molecular testing, the decision-making process has become complex. While established guidelines and pathways offer some guidance, they are difficult to utilize in a busy community practice and are not always implemented in the community. The rationale of the study was to identify a cohort of patients with lung adenocarcinoma at a City of Hope community site (n = 11) and utilize their case studies to develop a decision-making framework utilizing fast-and-frugal tree (FFT) heuristics. Most patients had stage IV (N = 9, 81.8%) disease at the time of the first consultation. The most common symptoms at initial presentation were cough (N = 5, 45.5%), shortness of breath (N = 3, 27.2%), and weight loss (N = 3, 27.2%). The Eastern Cooperative Oncology Group (ECOG) performance status ranged from 0-1 in all patients in this study. Distribution of molecular drivers among the patients were as follows: EGFR (N = 5, 45.5%), KRAS (N = 2, 18.2%), ALK (N = 2, 18.2%), MET (N = 2, 18.2%), and RET (N = 1, 9.1%). Seven initial FFTs were developed for the various case scenarios, but ultimately the decisions were condensed into one FFT, a molecular stage IV FFT, that arrived at accurate decisions without sacrificing initial information. While these FFT decision trees may seem arbitrary to an experienced oncologist at an academic site, the simplicity of their utility is essential for community practice where patients often do not get molecular testing and are not assigned proper therapy.

show abstract

Estimating Quantities: Comparing Simple Heuristics and Machine Learning Algorithms

Cited by 3 publications

References 11 publications

Integrating and testing natural frequencies, naïve Bayes, and fast-and-frugal trees.

Integrating and testing natural frequencies, naïve Bayes, and fast-and-frugal trees.

FFTrees: A toolbox to create, visualize, and evaluate fast-and-frugal decision trees

Complex Oncological Decision-Making Utilizing Fast-and-Frugal Trees in a Community Setting—Role of Academic and Hybrid Modeling

Contact Info

Product

Resources

About