MoËT: Mixture of Expert Trees and its application to verifiable reinforcement learning

Published in

Neural Networks(2019)

External Links:

Generate Graph DownloadPDF

TL;DR

MoËT, a novel model based on Mixture of Experts, consisting of decision tree experts and a generalized linear model gating function, which is more expressive than the standard decision tree and can be used in real-world supervised problems on which it outperforms other verifiable machine learning models.

Authors

Rishabh Singh

2 papers

Marko Vasic

1 papers

A. Petrović

1 papers

Kaiyuan Wang

1 papers

Mladen Nikolic

1 papers

S. Khurshid

1 papers

References62 items

Verified Probabilistic Policies for Deep Reinforcement Learning

Knowledge Distillation based Cooperative Reinforcement Learning for Connectivity Preservation in UAV Networks

XPM: An Explainable Deep Reinforcement Learning Framework for Portfolio Management

Explainable Deep Reinforcement Learning Using Introspection in a Non-episodic Task

Towards Scalable Verification of Deep Reinforcement Learning

MoËT: Mixture of Expert Trees and its application to verifiable reinforcement learning

Published in

Neural Networks(2019)

External Links:

Generate Graph DownloadPDF

TL;DR

Authors

Rishabh Singh

2 papers

Marko Vasic

1 papers

A. Petrović

1 papers

Kaiyuan Wang

1 papers

Mladen Nikolic

1 papers

S. Khurshid

1 papers

References62 items

Verified Probabilistic Policies for Deep Reinforcement Learning

Knowledge Distillation based Cooperative Reinforcement Learning for Connectivity Preservation in UAV Networks

XPM: An Explainable Deep Reinforcement Learning Framework for Portfolio Management

Explainable Deep Reinforcement Learning Using Introspection in a Non-episodic Task

Towards Scalable Verification of Deep Reinforcement Learning

Explainable AI and Reinforcement Learning—A Systematic Review of Current Approaches and Trends

KnowRU: Knowledge Reuse via Knowledge Distillation in Multi-Agent Reinforcement Learning

Diversity-driven knowledge distillation for financial trading using Deep Reinforcement Learning

Toward Verifiable and Privacy Preserving Machine Learning Prediction

Explainability in Deep Reinforcement Learning

Knowledge Distillation: A Survey

Explainable Reinforcement Learning: A Survey

A formal methods approach to interpretable reinforcement learning for robotic planning

Verifying Deep-RL-Driven Systems

Towards high-level, verifiable autonomous behaviors with temporal specifications

Imitation-Projected Programmatic Reinforcement Learning

Learning Causal State Representations of Partially Observable Environments

An inductive synthesis framework for verifiable reinforcement learning

Explainable Machine Learning for Scientific Insights and Discoveries

Formal Verification of Input-Output Mappings of Tree Ensembles

Hierarchical Routing Mixture of Experts

A guide to deep learning in healthcare

Deep learning for healthcare: review, opportunities and challenges

Formal Verification of Random Forests in Safety-Critical Applications

Learning Finite State Representations of Recurrent Policy Networks

Local Rule-Based Explanations of Black Box Decision Systems

Born Again Neural Networks

Verifiable Reinforcement Learning via Policy Extraction

A Survey of Methods for Explaining Black Box Models

Challenges in the Verification of Reinforcement Learning Algorithms

Training and Validating a Deep Convolutional Neural Network for Computer-Aided Detection and Classification of Abnormalities on Frontal Chest Radiographs

Towards A Rigorous Science of Interpretable Machine Learning

The mythos of model interpretability

Computer-Aided Diagnosis with Deep Learning Architecture: Applications to Breast Lesions in US Images and Pulmonary Nodules in CT Scans

“Why Should I Trust You?”: Explaining the Predictions of Any Classifier

Mastering the game of Go with deep neural networks and tree search

Dueling Network Architectures for Deep Reinforcement Learning

Policy Distillation

Distilling the Knowledge in a Neural Network

Human-level control through deep reinforcement learning

Soft decision trees

Twenty Years of Mixture of Experts

Decision trees: a recent overview

A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning

Notice of Retraction Review of decision trees

Z3: An Efficient SMT Solver

Model compression

Apprenticeship learning via inverse reinforcement learning

Is imitation learning the route to humanoid robots?

Scaling Up the Accuracy of Naive-Bayes Classifiers: A Decision-Tree Hybrid

Generalization in Reinforcement Learning: Successful Examples Using Sparse Coarse Coding

Adaptive Mixtures of Local Experts

Neuronlike adaptive elements that can solve difficult learning control problems

Large scale deep learning for computer aided detection of mammographic lesions

Uci machine learning repository

Handbook of Satis-ﬁability, Vol. 185 of Frontiers in Artiﬁcial Intelligence and Applications

Convergence Results for the Em Approach to Mixtures of Experts Architectures

Efficient memory-based learning for robot control

Noname manuscript No. (will be inserted by the editor) TEXPLORE: Real-Time Sample-Efficient Reinforcement Learning for Robots

We trained MoËT models for varying number of experts and their maximum depths. The number of experts used are

, Mo¨ET version with hard thresholding and softmax gating function which can be translated to an SMT formula amenable for veriﬁcation and is not hard to interpret in case of small models

Field of Study

Computer ScienceMathematicsMedicine

Journal Information

Name

Neural networks : the official journal of the International Neural Network Society

Page

85-117

Volume

Venue Information

Name

Neural Networks

Type

journal

URL

http://www.elsevier.com/locate/neunet

Alternate Names

Neural Netw

TL;DR

Authors

References62 items

Verified Probabilistic Policies for Deep Reinforcement Learning

Knowledge Distillation based Cooperative Reinforcement Learning for Connectivity Preservation in UAV Networks

XPM: An Explainable Deep Reinforcement Learning Framework for Portfolio Management

Explainable Deep Reinforcement Learning Using Introspection in a Non-episodic Task

Towards Scalable Verification of Deep Reinforcement Learning

TL;DR

Authors

References62 items

Verified Probabilistic Policies for Deep Reinforcement Learning

Knowledge Distillation based Cooperative Reinforcement Learning for Connectivity Preservation in UAV Networks

XPM: An Explainable Deep Reinforcement Learning Framework for Portfolio Management

Explainable Deep Reinforcement Learning Using Introspection in a Non-episodic Task

Towards Scalable Verification of Deep Reinforcement Learning

Explainable AI and Reinforcement Learning—A Systematic Review of Current Approaches and Trends

KnowRU: Knowledge Reuse via Knowledge Distillation in Multi-Agent Reinforcement Learning

Diversity-driven knowledge distillation for financial trading using Deep Reinforcement Learning

Toward Verifiable and Privacy Preserving Machine Learning Prediction

Explainability in Deep Reinforcement Learning

Knowledge Distillation: A Survey

Explainable Reinforcement Learning: A Survey

A formal methods approach to interpretable reinforcement learning for robotic planning

Verifying Deep-RL-Driven Systems

Towards high-level, verifiable autonomous behaviors with temporal specifications

Imitation-Projected Programmatic Reinforcement Learning

Learning Causal State Representations of Partially Observable Environments

An inductive synthesis framework for verifiable reinforcement learning

Explainable Machine Learning for Scientific Insights and Discoveries

Formal Verification of Input-Output Mappings of Tree Ensembles

Hierarchical Routing Mixture of Experts

A guide to deep learning in healthcare

Deep learning for healthcare: review, opportunities and challenges

Formal Verification of Random Forests in Safety-Critical Applications

Learning Finite State Representations of Recurrent Policy Networks

Local Rule-Based Explanations of Black Box Decision Systems

Born Again Neural Networks

Verifiable Reinforcement Learning via Policy Extraction

A Survey of Methods for Explaining Black Box Models

Challenges in the Verification of Reinforcement Learning Algorithms

Training and Validating a Deep Convolutional Neural Network for Computer-Aided Detection and Classification of Abnormalities on Frontal Chest Radiographs

Towards A Rigorous Science of Interpretable Machine Learning

The mythos of model interpretability

Computer-Aided Diagnosis with Deep Learning Architecture: Applications to Breast Lesions in US Images and Pulmonary Nodules in CT Scans

“Why Should I Trust You?”: Explaining the Predictions of Any Classifier

Mastering the game of Go with deep neural networks and tree search

Dueling Network Architectures for Deep Reinforcement Learning

Policy Distillation

Distilling the Knowledge in a Neural Network

Human-level control through deep reinforcement learning

Soft decision trees

Twenty Years of Mixture of Experts

Decision trees: a recent overview

A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning

Notice of Retraction Review of decision trees

Z3: An Efficient SMT Solver

Model compression

Apprenticeship learning via inverse reinforcement learning

Is imitation learning the route to humanoid robots?

Scaling Up the Accuracy of Naive-Bayes Classifiers: A Decision-Tree Hybrid

Generalization in Reinforcement Learning: Successful Examples Using Sparse Coarse Coding

Adaptive Mixtures of Local Experts

Neuronlike adaptive elements that can solve difficult learning control problems

Large scale deep learning for computer aided detection of mammographic lesions

Uci machine learning repository

Handbook of Satis-ﬁability, Vol. 185 of Frontiers in Artiﬁcial Intelligence and Applications

BORN AGAIN TREES

Convergence Results for the Em Approach to Mixtures of Experts Architectures

Efficient memory-based learning for robot control

Noname manuscript No. (will be inserted by the editor) TEXPLORE: Real-Time Sample-Efficient Reinforcement Learning for Robots

We trained MoËT models for varying number of experts and their maximum depths. The number of experts used are

, Mo¨ET version with hard thresholding and softmax gating function which can be translated to an SMT formula amenable for veriﬁcation and is not hard to interpret in case of small models

Field of Study

Journal Information

Name

Page

Volume

Venue Information

Name