Multiple Instance Learning with Trainable Decision Tree Ensembles

Andrei V. Konstantinov; Lev V. Utkin

RECORD

Multiple Instance Learning with Trainable Decision Tree Ensembles

Andrei V. Konstantinov, Lev V. Utkin

Machine Learning · Submitted 13 Feb 2023

abstract compiled · 2026-09-23 · no language-model review

The authors' main claim Numerical experiments with tabular datasets illustrate STE-MIL.

how it is stated · asserted 0.99 read the record →

Comment (0) Read on arXiv ↗

Cite ▾

Claim paper

The compiled record

The paper's own sentences, arranged. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.

Open the compiled record

Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-09-23: 1 request, 5,460 input tokens.

What the abstract itself states, as probabilities: claims first 0.09 · states limitation 0.28 · human participants 0.04 · controlled comparison 0.07 · randomized assignment 0.02 · states quantitative result 0.04 · compares with prior results 0.10 · mentions code or data availability 0.98.

The abstract, sentence by sentence

arXiv listing abstract v1. Results and the main claim are underlined; show every role.

A new random forest based model for solving the Multiple Instance Learning (MIL) problem under small tabular data, called Soft Tree Ensemble MIL (STE-MIL), is proposed. A new type of soft decision trees is considered, which is similar to the well-known soft oblique trees, but with a smaller number of trainable parameters. In order to train the trees, it is proposed to convert them into neural networks of a specific form, which approximate the tree functions. It is also proposed to aggregate the instance and bag embeddings (output vectors) by using the attention mechanism. The whole STE-MIL model, including soft decision trees, neural networks, the attention mechanism and a classifier, is trained in an end-to-end manner. Numerical experiments with tabular datasets illustrate STE-MIL. The corresponding code implementing the model is publicly available.

background contribution method result interpretation main claim

What the classifier reports for each sentence. The number is the probability it gave the label.
#	sentence	role	the authors' own work	states the main conclusion	strength
s0	A new random forest based model for solving the Multiple Instance Learning (MIL) problem under small tabular data, called Soft Tree Ensemble MIL (STE-MIL), is proposed.	contribution 0.99	0.89	0.19	asserted 1.00
s1	A new type of soft decision trees is considered, which is similar to the well-known soft oblique trees, but with a smaller number of trainable parameters.	method 0.62	0.35	0.03	asserted 1.00
s2	In order to train the trees, it is proposed to convert them into neural networks of a specific form, which approximate the tree functions.	method 0.93	0.83	0.03	asserted 0.97
s3	It is also proposed to aggregate the instance and bag embeddings (output vectors) by using the attention mechanism.	method 0.76	0.84	0.04	asserted 0.98
s4	The whole STE-MIL model, including soft decision trees, neural networks, the attention mechanism and a classifier, is trained in an end-to-end manner.	method 0.96	0.86	0.08	asserted 1.00
s5	Numerical experiments with tabular datasets illustrate STE-MIL.	method 0.62	0.91	0.09	asserted 0.99
s6	The corresponding code implementing the model is publicly available.	contribution 0.89	0.72	0.03	asserted 1.00

What kind of paper the classifier reports this is: method or tool 1.00 · evidence from observation or experiment 0.68.

What was examined, and what was not

sources read

arXiv listing abstract v1

not examined

full text
figures and tables
references and citing sentences
relations to other papers

levels run

P abstract

not run

S sections
T body sentences
C citations
R cross-paper relations

Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. Every number in the measurements table is the authors', checked by code. Record as JSON

The paper

Original abstract

A new random forest based model for solving the Multiple Instance Learning (MIL) problem under small tabular data, called Soft Tree Ensemble MIL (STE-MIL), is proposed. A new type of soft decision trees is considered, which is similar to the well-known soft oblique trees, but with a smaller number of trainable parameters. In order to train the trees, it is proposed to convert them into neural networks of a specific form, which approximate the tree functions. It is also proposed to aggregate the instance and bag embeddings (output vectors) by using the attention mechanism. The whole STE-MIL model, including soft decision trees, neural networks, the attention mechanism and a classifier, is trained in an end-to-end manner. Numerical experiments with tabular datasets illustrate STE-MIL. The corresponding code implementing the model is publicly available.

Human review

How Pith checks papers → · Full audit of this paper →

How the literature uses it

What this paper does with the arXiv papers it cites

Each citing sentence is taken from the citing paper's LaTeX source and typed by Pith's classifier. A label appears at probability 0.8 or more (0.95 for support), a contest at 0.7 or more; the quoted words are the citing authors'. How each label was checked · JSON

What it cites · 19 sentences without a confident label

Uses the method or tool of 1 paper

[equation] where the indicator functions are replaced with the so-called $σ$-hard indicator approximations [ref] which apply quantized functions in the forward pass, but uses a smooth activation function in the backward pass to propagate.

Learning Accurate Decision Trees with Bandit Feedback via Quantized Gradient Descent · Oblique binary soft trees · 0.83

Cites as background 3 papers

Madaan et al. [ref] presented dense gradient trees and an transformer based on the trees, which is called Treeformer.

Treeformer: Dense Gradient Trees for Efficient Attention Computation · Related work · 0.94
Frosst and Hinton [ref] take the knowledge acquired by a neural net and express the same knowledge in a model that relies on hierarchical decisions instead, explaining a particular decision would be much easier.

Distilling a Neural Network Into a Soft Decision Tree · Related work · 0.89
[ref] show how neural models can be used to realize piece-wise constant functions such as decision trees.

Oblique Decision Trees from Derivatives of ReLU Networks · Related work · 0.86

Discussion

Know something about this paper? Leave a comment, or highlight any passage above to comment on it.