Robot hand-eye cooperation based on improved inverse reinforcement learning

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Ning Yu ◽  
Lin Nan ◽  
Tao Ku
Keyword(s):  
Decision Making ◽  
Reinforcement Learning ◽  
Visual Information ◽  
Industrial Robots ◽  
Robot Hand ◽  
Inverse Reinforcement Learning ◽  
Generative Adversarial Network ◽  
Content Type ◽  
Coordination Model ◽  
Proposed Model

Purpose How to make accurate action decisions based on visual information is one of the important research directions of industrial robots. The purpose of this paper is to design a highly optimized hand-eye coordination model of the robot to improve the robots’ on-site decision-making ability. Design/methodology/approach The combination of inverse reinforcement learning (IRL) algorithm and generative adversarial network can effectively reduce the dependence on expert samples and robots can obtain the decision-making performance that the degree of optimization is not lower than or even higher than that of expert samples. Findings The performance of the proposed model is verified in the simulation environment and real scene. By monitoring the reward distribution of the reward function and the trajectory of the robot, the proposed model is compared with other existing methods. The experimental results show that the proposed model has better decision-making performance in the case of less expert data. Originality/value A robot hand-eye cooperation model based on improved IRL is proposed and verified. Empirical investigations on real experiments reveal that overall, the proposed approach tends to improve the real efficiency by more than 10% when compared to alternative hand-eye cooperation methods.

scholarly journals Deep inverse reinforcement learning for structural evolution of small molecules

2020 ◽  
Author(s):  
Brighter Agyemang ◽  
Wei-Ping Wu ◽  
Daniel Addo ◽  
Michael Y Kpiebaareh ◽  
Ebenezer Nanor ◽  
...  
Keyword(s):  
Reinforcement Learning ◽  
High Throughput Screening ◽  
Structural Evolution ◽  
Search Space ◽  
New Drugs ◽  
Inverse Reinforcement Learning ◽  
Generative Adversarial Network ◽  
Entropy Maximization ◽  
Reward Function ◽  
Adversarial Network

Abstract The size and quality of chemical libraries to the drug discovery pipeline are crucial for developing new drugs or repurposing existing drugs. Existing techniques such as combinatorial organic synthesis and high-throughput screening usually make the process extraordinarily tough and complicated since the search space of synthetically feasible drugs is exorbitantly huge. While reinforcement learning has been mostly exploited in the literature for generating novel compounds, the requirement of designing a reward function that succinctly represents the learning objective could prove daunting in certain complex domains. Generative adversarial network-based methods also mostly discard the discriminator after training and could be hard to train. In this study, we propose a framework for training a compound generator and learn a transferable reward function based on the entropy maximization inverse reinforcement learning (IRL) paradigm. We show from our experiments that the IRL route offers a rational alternative for generating chemical compounds in domains where reward function engineering may be less appealing or impossible while data exhibiting the desired objective is readily available.

scholarly journals The impact of product presentation on decision-making and purchasing

2019 ◽  
Vol 22 (3) ◽  
pp. 365-380 ◽  
Author(s):  
Rosy Boardman ◽  
Helen McCormick
Keyword(s):  
Decision Making ◽  
Eye Tracking ◽  
Visual Information ◽  
Age Groups ◽  
Target Market ◽  
Consumer Decision Making ◽  
Consumer Decision ◽  
Content Type ◽  
Product Presentation ◽  
The Impact

Purpose This paper aims to investigate how apparel product presentation influences consumer decision-making and whether there are any differences between age groups. Design/methodology/approach A mixed methodology was used including eye-tracking and qualitative in-depth interviews, with a purposive sample of 50 participants between age 20 and 70. Findings A higher number of product presentation features resulted in increased positive visual, cognitive and affective responses as consumers wanted as much visual information as possible to aid decision-making. Images of models attracted the most attention and were the most influential product presentation feature, followed by mannequin images and the zoom function. The 20 s spent much less time viewing and interacting with the product presentation features than middle age groups (30 s-50 s), had minimal fixations on mannequin images and had a much quicker decision-making process than other age groups. Practical implications The research informs retailers which product presentation features are the most effective for their target market to aid consumer decision-making with the aim of reducing returns. Originality/value The paper contributes to the literature by providing more in-depth insights than previous studies into the impact of online product presentation on consumer decision-making by using qualitative research and eye-tracking. The research also explores more product presentation features than previous research and investigates the presentation of apparel products, which are notoriously the most difficult products for consumers to assess online. The research is unique in its exploration of age differences in relation to product presentation features.

Sustainable efficiency assessment of private diagnostic laboratories under uncertainty

2020 ◽  
Vol 15 (3) ◽  
pp. 1069-1103
Author(s):  
Niloufar Ghafari Someh ◽  
Mir Saman Pishvaee ◽  
Seyed Jafar Sadjadi ◽  
Roya Soltani
Keyword(s):  
Decision Making ◽  
Sustainable Development ◽  
Managerial Decision ◽  
Diagnostic Laboratory ◽  
Content Type ◽  
Development Indicators ◽  
Sustainable Development Indicators ◽  
Proposed Model ◽  
Decision Making Unit ◽  
Medical Diagnostic

Purpose Assessing the performance of medical laboratories plays an important role in the quality of health services. However, because of imprecise data, reliable results from laboratory performance cannot be obtained easily. The purpose of this paper is to illustrate the use of interval network data envelopment analysis (INDEA) based on sustainable development indicators under uncertainty. Design/methodology/approach In this study, each medical diagnostic laboratory is considered as a decision-making unit (DMU) and an INDEA model is used for calculating the efficiency of each medical diagnostic laboratory under imprecise inputs and outputs. The proposed model helps provide managers with effective performance scores for deficiencies and business improvements. The proposed model with realistic efficiency scores can help administrators manage their deficiencies and ultimately improve their business. Findings The results indicate that uncertainty can lead to changes in performance scores, rankings and performance classifications. Therefore, the use of DEA models under certainty can be potentially misleading. Originality/value The contribution of this study provides useful insights into the use of INDEA as a modeling tool to aid managerial decision-making in assessing efficiency of medical diagnostic laboratories based on sustainable development indicators under uncertainty.

An estimation model of construction project segmentation for optimum project pricing

2020 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Fang-Jye Shiue ◽  
Hsin-Yun Lee ◽  
Meng-Cong Zheng ◽  
Akhmad F.K. Khitam ◽  
Sintayehu Assefa
Keyword(s):  
Decision Making ◽  
Construction Projects ◽  
Critical Issue ◽  
Multiple Attribute Decision Making ◽  
Estimation Model ◽  
Content Type ◽  
Proposed Model ◽  
Multiple Attribute ◽  
Bid Price ◽  
Construction Bids

PurposeFor large projects, project segmentation and planning the size of contract packages in construction bids is a complex and critical issue. Due to the nature of construction projects, which frequently have large budgets, long durations and many activities with complex procedures, project segmentation involves complicated decision-making. To fill this gap, this study aims to develop an integrated model for planning project segmentation.Design/methodology/approachThe proposed model integrates a simulation and multiple attribute decision-making method. The simulation is used to evaluate the bidding outcome of various project segmentations. The owner can then determine the bid-price behavior of contractors in response to varying work package sizes. The multiple attribute decision-making method is used to select the optimal segmentation solution from the simulated scenarios.FindingsThe proposed model is applied to a large road preservation project in Indonesia and incorporates bid participants and market conditions. The model provides seven scenarios for segmentation. The range of scenarios captures increasing competitiveness in the construction with the average bid price becoming gradually more beneficial for the owner. The model also utilizes a multiple attribute decision-making method to select the optimum scenario for the owner.Originality/valueThis study presents an applicable model for project segmentation that is useful for both project owners and contractors. By utilizing the proposed model, a project owner can segment a large project into smaller contract packages to create improved project pricing.

Grey target decision-making model of interval grey number based on cone volume

2017 ◽  
Vol 7 (2) ◽  
pp. 247-258 ◽  
Author(s):  
Lizhen Wang ◽  
Wuyong Qian
Keyword(s):  
Decision Making ◽  
Entropy Method ◽  
Maximal Entropy ◽  
Information Distortion ◽  
Content Type ◽  
Model Based ◽  
Proposed Model ◽  
Grey Number ◽  
Decision Making Model ◽  
The Impact

Purpose The purpose of this paper is to propose a grey target decision model based on cobweb area in order to overcome the effect and influence from the extreme value of the index on the decision result. However, it does not take into account the impact of the correlation between indicators on the angle of the index, and produce a certain degree decision information distortion as a result of the equal angle between the indicators. In order to solve the above problems, a novel grey decision-making model based on cone volume is proposed. Design/methodology/approach In this paper, the model uses the whitening weight function to whiten the interval grey number, and the Delphi method and the maximal entropy method are exploited to integrate the weight of the index. On the basis of this, the center of the bull’s eye, the weight and the index value are constructed as the center circle, the radius, and the high cone, respectively. The scheme is selected by the volume of the cone, the decision is made according to the order relation, and the example is utilized to prove and analyze the validity of the proposed model. Findings The results show that the proposed model can well improve the traditional grey target decision-making model from the modeling object and modeling method. Practical implications The method exposed in the paper can be used to deal with the grey target decision-making problems which characteristics are multi-indexes, and the attribute values are interval grey numbers. Originality/value The paper succeeds in overcoming the disadvantages of grey target decision making based on the target center distance and the cobweb area.

A novel model for risk management of outsourced construction projects using decision-making methods: a case study

2020 ◽  
Vol 10 (2) ◽  
pp. 97-123 ◽  
Author(s):  
Amin Mahmoudi ◽  
Mehdi Abbasi ◽  
Xiaopeng Deng ◽  
Muhammad Ikram ◽  
Salman Yeganeh
Keyword(s):  
Decision Making ◽  
Construction Projects ◽  
Project Managers ◽  
Analytic Hierarchy ◽  
Suitable Alternative ◽  
Content Type ◽  
Proposed Model ◽  
The Cost ◽  
First Time

PurposeSelecting a suitable contract to outsource construction projects is an ongoing concern for project managers and organizational directors. This study aims to propose a comprehensive model to manage the risks of outsourced construction project contracts.Design/methodology/approachTo employ the proposed model, firstly, the types of contracts and risks in the organization should be identified, then, to prioritize the contracts, the identified risks are considered as criteria. After receiving the experts' opinions, the best–worst method (BWM) integrated with grey relation analysis (GRA) method was used to prioritize the contracts. BWM and GRA are multi-criteria decision-making methods with different approaches and applications. In the current study, BWM has been employed to calculate the weights of criteria because it has better performance than other methods such as the analytic hierarchy process (AHP). After calculating the weights of criteria, the GRA method has been utilized for ranking the alternatives.FindingsAccording to the results obtained from the case study, the cost plus award fee contract is the most suitable alternative for outsourcing construction projects. The proposed methodology can be practically applied through different types of the projects such as construction or “engineering, procurement and construction”.Originality/valueTo the best of our knowledge, this is the first time a conceptual model has been proposed to select an appropriate contract for construction projects. Also, for the first time, the BWM integrated with GRA method has been used to prioritize project contracts based on the potential risks. The proposed model can contribute to project managers for selecting a suitable contract with the least risk in construction projects.

Probabilistic Approach to Robot Group Control

2011 ◽  
Vol 317-319 ◽  
pp. 742-749 ◽  
Author(s):  
Tomislav Stipančić ◽  
Bojan Jerbić ◽  
Petar Ćurković
Keyword(s):  
Decision Making ◽  
Probabilistic Approach ◽  
Industrial Applications ◽  
Industrial Robots ◽  
Working Environment ◽  
Human Intervention ◽  
Specific Domain ◽  
Proposed Model ◽  
Group Control ◽  
Robot Group

The objective of this paper is to discuss the probabilistic part of the model for robot group control applied in industrial applications. The proposed model is based on well-known concepts of Ubiquitous Computing [1] and enables contextual perception of a working environment. Compared with classical industrial robots, usually preprogrammed for a limited number of operations / actions, the system based on this model can react in uncertain situations and scenarios. The model combines ontology to describe the specific domain of interest and decision–making mechanisms based on Bayesian Networks (BN) to enable the work of a single robot without human intervention by learning Behavioral Patterns (BP) of other robots in the group.

scholarly journals Analysis and Validation of Cross-Modal Generative Adversarial Network for Sensory Substitution

2021 ◽  
Vol 18 (12) ◽  
pp. 6216
Author(s):  
Mooseop Kim ◽  
YunKyung Park ◽  
KyeongDeok Moon ◽  
Chi Yoon Jeong
Keyword(s):  
Visual Information ◽  
Experimental Results ◽  
Auditory Sensitivity ◽  
Sensory Substitution ◽  
Sensitivity Analyses ◽  
Auditory Signal ◽  
Behavioral Experiments ◽  
Generative Adversarial Network ◽  
Adversarial Network ◽  
Proposed Model

Visual-auditory sensory substitution has demonstrated great potential to help visually impaired and blind groups to recognize objects and to perform basic navigational tasks. However, the high latency between visual information acquisition and auditory transduction may contribute to the lack of the successful adoption of such aid technologies in the blind community; thus far, substitution methods have remained only laboratory-scale research or pilot demonstrations. This high latency for data conversion leads to challenges in perceiving fast-moving objects or rapid environmental changes. To reduce this latency, prior analysis of auditory sensitivity is necessary. However, existing auditory sensitivity analyses are subjective because they were conducted using human behavioral analysis. Therefore, in this study, we propose a cross-modal generative adversarial network-based evaluation method to find an optimal auditory sensitivity to reduce transmission latency in visual-auditory sensory substitution, which is related to the perception of visual information. We further conducted a human-based assessment to evaluate the effectiveness of the proposed model-based analysis in human behavioral experiments. We conducted experiments with three participant groups, including sighted users (SU), congenitally blind (CB) and late-blind (LB) individuals. Experimental results from the proposed model showed that the temporal length of the auditory signal for sensory substitution could be reduced by 50%. This result indicates the possibility of improving the performance of the conventional vOICe method by up to two times. We confirmed that our experimental results are consistent with human assessment through behavioral experiments. Analyzing auditory sensitivity with deep learning models has the potential to improve the efficiency of sensory substitution.

Research on decision-making of autonomous vehicle following based on reinforcement learning method

2019 ◽  
Vol 46 (3) ◽  
pp. 444-452 ◽  
Author(s):  
Hongbo Gao ◽  
Guanya Shi ◽  
Kelong Wang ◽  
Guotao Xie ◽  
Yuchao Liu
Keyword(s):  
Decision Making ◽  
Reinforcement Learning ◽  
Autonomous Vehicles ◽  
Autonomous Vehicle ◽  
Learning System ◽  
Explicit Calculation ◽  
Gradient Descent Method ◽  
Learning Method ◽  
Content Type ◽  
Car Following

Purpose Over the past decades, there has been significant research effort dedicated to the development of autonomous vehicles. The decision-making system, which is responsible for driving safety, is one of the most important technologies for autonomous vehicles. The purpose of this study is the use of an intensive learning method combined with car-following data by a driving simulator to obtain an explanatory learning following algorithm and establish an anthropomorphic car-following model. Design/methodology/approach This paper proposed car-following method based on reinforcement learning for autonomous vehicles decision-making. An approximator is used to approximate the value function by determining state space, action space and state transition relationship. A gradient descent method is used to solve the parameter. Findings The effect of car-following on certain driving styles is initially achieved through the simulation of step conditions. The effect of car-following initially proves that the reinforcement learning system is more adaptive to car following and that it has certain explanatory and stability based on the explicit calculation of R. Originality/value The simulation results show that the car-following method based on reinforcement learning for autonomous vehicle decision-making realizes reliable car-following decision-making and has the advantages of simple sample, small amount of data, simple algorithm and good robustness.

Goal programming optimization model under uncertainty and the critical areas characterization in humanitarian logistics management

2019 ◽  
Vol 9 (1) ◽  
pp. 82-107 ◽  
Author(s):  
Mario Chong ◽  
Juan G. Lazo Lazo ◽  
Maria Cristina Pereda ◽  
Juan Manuel Machuca De Pina
Keyword(s):  
Decision Making ◽  
Disaster Management ◽  
Low Income ◽  
Point Of View ◽  
Decision Makers ◽  
Humanitarian Crisis ◽  
Content Type ◽  
Critical Areas ◽  
Proposed Model ◽  
Wide Range

Purpose The purpose of this paper is to improve disaster management models, have an optimal distribution of assets, reduce human suffering in a crisis and find a good solution for warehouse locations, distribution points, inventory levels and costs, considering the uncertainty of a wide range of variables, to serve as a support model for decision making in real situations. Design/methodology/approach A model is developed based on the recent models. It includes structured and non-structured data (historical knowledge) from a humanitarian perspective. This model considers the uncertainty in a landslide and flood area and it is applied in a representative Peruvian city. Findings The proposed model can be used to determine humanitarian aid supply and its distribution with uncertainty, regarding the affected population and its resilience. This model presents a different point of view from the efficiency of the logistics perspective, to identify the level of trust between all the stakeholders (public, private and academic). The finding provides a new insight in disaster management to cover the gap between applied research and human behavior in crisis. Research limitations/implications In this study the access of reliable information is limited. Practical implications This paper provides an operation model with uncertainty in a humanitarian crisis and a decision-making tool with some recommendation for further public policies. Originality/value This study presents a model for decision makers in a low-income zone and highlights the importance of preparedness in the humanitarian system. This paper expands the discussion of how the mathematical models and human behaviors interact with different perspectives in a humanitarian crisis.

Sign in / Sign up

Export Citation Format

Share Document