scholarly journals Purely Declarative Action Descriptions are Overrated: Classical Planning with Simulators

2017 ◽  
Author(s):  
Guillem Francès ◽  
Miquel Ramírez ◽  
Nir Lipovetzky ◽  
Hector Geffner
Keyword(s):  
State Of The Art ◽  
Black Box ◽  
Initial State ◽  
Planning Algorithms ◽  
Control Knowledge ◽  
Domain Independent ◽  
Action Structure

Classical planning is concerned with problems where a goal needs to be reached from a known initial state by doing actions with deterministic, known effects. Classical planners, however, deal only with classical problems that can be expressed in declarative planning languages such as STRIPS or PDDL. This prevents their use on problems that are not easy to model declaratively or whose dynamics are given via simulations. Simulators do not provide a declarative representation of actions, but simply return successor states. The question we address in this paper is: can a planner that has access to the structure of states and goals only, approach the performance of planners that also have access to the structure of actions expressed in PDDL? To answer this, we develop domain-independent, black box planning algorithms that completely ignore action structure, and show that they match the performance of state-of-the-art classical planners on the standard planning benchmarks. Effective black box algorithms open up new possibilities for modeling and for expressing control knowledge, which we also illustrate.

scholarly journals Optimizing Discrete Spaces via Expensive Evaluations: A Learning to Search Framework

2020 ◽  
Vol 34 (04) ◽  
pp. 3773-3780 ◽  
Author(s):  
Aryan Deshwal ◽  
Syrine Belakaria ◽  
Janardhan Rao Doppa ◽  
Alan Fern
Keyword(s):  
Optimization Problems ◽  
State Of The Art ◽  
Black Box ◽  
Combinatorial Structures ◽  
High Performing ◽  
New Learning ◽  
Local Search Procedure ◽  
Complex Optimization ◽  
Control Knowledge ◽  
Discrete Spaces

We consider the problem of optimizing expensive black-box functions over discrete spaces (e.g., sets, sequences, graphs). The key challenge is to select a sequence of combinatorial structures to evaluate, in order to identify high-performing structures as quickly as possible. Our main contribution is to introduce and evaluate a new learning-to-search framework for this problem called L2S-DISCO. The key insight is to employ search procedures guided by control knowledge at each step to select the next structure and to improve the control knowledge as new function evaluations are observed. We provide a concrete instantiation of L2S-DISCO for local search procedure and empirically evaluate it on diverse real-world benchmarks. Results show the efficacy of L2S-DISCO over state-of-the-art algorithms in solving complex optimization problems.

Review of the Applications of Deep Learning in Bioinformatics

2021 ◽  
Vol 15 (8) ◽  
pp. 898-911
Author(s):  
Yongqing Zhang ◽  
Jianrong Yan ◽  
Siyu Chen ◽  
Meiqin Gong ◽  
Dongrui Gao ◽  
...  
Keyword(s):  
Deep Learning ◽  
Drug Discovery ◽  
Biomedical Imaging ◽  
State Of The Art ◽  
Black Box ◽  
Medical Data ◽  
Biological Data ◽  
High Dimensional ◽  
Biological Research ◽  
Process Data

Rapid advances in biological research over recent years have significantly enriched biological and medical data resources. Deep learning-based techniques have been successfully utilized to process data in this field, and they have exhibited state-of-the-art performances even on high-dimensional, nonstructural, and black-box biological data. The aim of the current study is to provide an overview of the deep learning-based techniques used in biology and medicine and their state-of-the-art applications. In particular, we introduce the fundamentals of deep learning and then review the success of applying such methods to bioinformatics, biomedical imaging, biomedicine, and drug discovery. We also discuss the challenges and limitations of this field, and outline possible directions for further research.

scholarly journals Efficient Lifted Planning with Regression-Based Heuristics

2021 ◽  
Author(s):  
Hadi Qovaizi
Keyword(s):  
Organic Synthesis ◽  
State Of The Art ◽  
Transition System ◽  
Modern State ◽  
Synthetic Approach ◽  
Best First Search ◽  
Planning Task ◽  
International Planning ◽  
Domain Independent ◽  
Novel Algorithm

Modern state-of-the-art planners operate by generating a grounded transition system prior to performing search for a solution to a given planning task. Some tasks involve a significant number of objects or entail managing predicates and action schemas with a significant number of arguments. Hence, this instantiation procedure can exhaust all available memory and therefore prevent a planner from performing search to find a solution. This thesis explores this limitation by presenting a benchmark set of problems based on Organic Chemistry Synthesis that was submitted to the latest International Planning Competition (IPC-2018). This benchmark was constructed to gauge the performance of the competing planners given that instantiation is an issue. Furthermore, a novel algorithm, the Regression-Based Heuristic Planner (RBHP), is developed with the aim of averting this issue. RBHP was inspired by the retro-synthetic approach commonly used to solve organic synthesis problems efficiently. RBHP solves planning tasks by applying domain independent heuristics, computed by regression, and performing best-first search. In contrast to most modern planners, RBHP computes heuristics backwards by applying the goal-directed regression operator. However, the best-first search proceeds forward similar to other planners. The proposed planner is evaluated on a set of planning tasks included in previous International Planning Competitions (IPC) against a subset of the top scoring state-of-the-art planners submitted to the IPC-2018.

scholarly journals The Audio Auditor: User-Level Membership Inference in Internet of Things Voice Services

2021 ◽  
Vol 2021 (1) ◽  
pp. 209-228
Author(s):  
Yuantian Miao ◽  
Minhui Xue ◽  
Chao Chen ◽  
Lei Pan ◽  
Jun Zhang ◽  
...  
Keyword(s):  
Internet Of Things ◽  
State Of The Art ◽  
Rapid Development ◽  
Black Box ◽  
Problem Space ◽  
Specific Data ◽  
Learning Techniques ◽  
Audio Data ◽  
Iot Devices ◽  
Asr System

AbstractWith the rapid development of deep learning techniques, the popularity of voice services implemented on various Internet of Things (IoT) devices is ever increasing. In this paper, we examine user-level membership inference in the problem space of voice services, by designing an audio auditor to verify whether a specific user had unwillingly contributed audio used to train an automatic speech recognition (ASR) model under strict black-box access. With user representation of the input audio data and their corresponding translated text, our trained auditor is effective in user-level audit. We also observe that the auditor trained on specific data can be generalized well regardless of the ASR model architecture. We validate the auditor on ASR models trained with LSTM, RNNs, and GRU algorithms on two state-of-the-art pipelines, the hybrid ASR system and the end-to-end ASR system. Finally, we conduct a real-world trial of our auditor on iPhone Siri, achieving an overall accuracy exceeding 80%. We hope the methodology developed in this paper and findings can inform privacy advocates to overhaul IoT privacy.

scholarly journals Decision-Theoretic Planning with non-Markovian Rewards

2006 ◽  
Vol 25 ◽  
pp. 17-74 ◽  
Author(s):  
S. Thiebaux ◽  
C. Gretton ◽  
J. Slaney ◽  
D. Price ◽  
F. Kabanza
Keyword(s):  
Decision Process ◽  
Single Interface ◽  
Reward Function ◽  
Domain Specific ◽  
Current State ◽  
Markovian Decision Process ◽  
Solution Methods ◽  
Control Knowledge ◽  
Search Control ◽  
Domain Independent

A decision process in which rewards depend on history rather than merely on the current state is called a decision process with non-Markovian rewards (NMRDP). In decision-theoretic planning, where many desirable behaviours are more naturally expressed as properties of execution sequences rather than as properties of states, NMRDPs form a more natural model than the commonly adopted fully Markovian decision process (MDP) model. While the more tractable solution methods developed for MDPs do not directly apply in the presence of non-Markovian rewards, a number of solution methods for NMRDPs have been proposed in the literature. These all exploit a compact specification of the non-Markovian reward function in temporal logic, to automatically translate the NMRDP into an equivalent MDP which is solved using efficient MDP solution methods. This paper presents NMRDPP (Non-Markovian Reward Decision Process Planner), a software platform for the development and experimentation of methods for decision-theoretic planning with non-Markovian rewards. The current version of NMRDPP implements, under a single interface, a family of methods based on existing as well as new approaches which we describe in detail. These include dynamic programming, heuristic search, and structured methods. Using NMRDPP, we compare the methods and identify certain problem features that affect their performance. NMRDPP's treatment of non-Markovian rewards is inspired by the treatment of domain-specific search control knowledge in the TLPlan planner, which it incorporates as a special case. In the First International Probabilistic Planning Competition, NMRDPP was able to compete and perform well in both the domain-independent and hand-coded tracks, using search control knowledge in the latter.

Interlaced Attitude Estimation for Spacecraft Using Vector Observations

2021 ◽  
Vol 01 (03) ◽  
Author(s):  
Lubin Chang
Keyword(s):  
Prior Information ◽  
State Of The Art ◽  
Estimation Method ◽  
Attitude Estimation ◽  
The Body ◽  
Body Frame ◽  
Initial Value ◽  
Initial State ◽  
Initial Errors ◽  
Initial Attitude

This paper proposes an interlaced attitude estimation method for spacecraft using vector observations, which can simultaneously estimate the constant attitude at the very start and the attitude of the body frame relative to its initial state. The arbitrary initial attitude, described by constant attitude at the very start, is determined using quaternion estimator which requires no prior information. The multiplicative extended Kalman filter (EKF) is competent for estimating the attitude of the body frame relative to its initial state since the initial value of this attitude is exactly known. The simulation results show that the proposed algorithms could achieve better performance compared with the state-of-the-art algorithms even with extreme large initial errors. Meanwhile, the computational burden is also much less than that of the advanced nonlinear attitude estimators.

Generating New Space-Filling Test Instances for Continuous Black-Box Optimization

2020 ◽  
Vol 28 (3) ◽  
pp. 379-404
Author(s):  
Mario A. Muñoz ◽  
Kate Smith-Miles
Keyword(s):  
Feature Vector ◽  
State Of The Art ◽  
Evolutionary Process ◽  
Black Box ◽  
Two Dimensional ◽  
Entire Space ◽  
Test Functions ◽  
New Space ◽  
Instance Space ◽  
Different Characteristics

This article presents a method to generate diverse and challenging new test instances for continuous black-box optimization. Each instance is represented as a feature vector of exploratory landscape analysis measures. By projecting the features into a two-dimensional instance space, the location of existing test instances can be visualized, and their similarities and differences revealed. New instances are generated through genetic programming which evolves functions with controllable characteristics. Convergence to selected target points in the instance space is used to drive the evolutionary process, such that the new instances span the entire space more comprehensively. We demonstrate the method by generating two-dimensional functions to visualize its success, and ten-dimensional functions to test its scalability. We show that the method can recreate existing test functions when target points are co-located with existing functions, and can generate new functions with entirely different characteristics when target points are located in empty regions of the instance space. Moreover, we test the effectiveness of three state-of-the-art algorithms on the new set of instances. The results demonstrate that the new set is not only more diverse than a well-known benchmark set, but also more challenging for the tested algorithms. Hence, the method opens up a new avenue for developing test instances with controllable characteristics, necessary to expose the strengths and weaknesses of algorithms, and drive algorithm development.

Sign in / Sign up

Export Citation Format

Share Document