scholarly journals Mixture of Experts with Entropic Regularization for Data Classification

Entropy ◽  
2019 ◽  
Vol 21 (2) ◽  
pp. 190
Author(s):  
Billy Peralta ◽  
Ariel Saavedra ◽  
Luis Caro ◽  
Alvaro Soto
Keyword(s):  
Weather Forecasting ◽  
Complex Data ◽  
Mixture Of Experts ◽  
Proposed Model ◽  
Entropic Regularization ◽  
Winner Takes All ◽  
Multiple Experts ◽  
Gating Network ◽  
The Cost ◽  
Future Work

Today, there is growing interest in the automatic classification of a variety of tasks, such as weather forecasting, product recommendations, intrusion detection, and people recognition. “Mixture-of-experts” is a well-known classification technique; it is a probabilistic model consisting of local expert classifiers weighted by a gate network that is typically based on softmax functions, combined with learnable complex patterns in data. In this scheme, one data point is influenced by only one expert; as a result, the training process can be misguided in real datasets for which complex data need to be explained by multiple experts. In this work, we propose a variant of the regular mixture-of-experts model. In the proposed model, the cost classification is penalized by the Shannon entropy of the gating network in order to avoid a “winner-takes-all” output for the gating network. Experiments show the advantage of our approach using several real datasets, with improvements in mean accuracy of 3–6% in some datasets. In future work, we plan to embed feature selection into this model.

scholarly journals APPLICATION OF ECONOMIC AND MATHEMATICAL METHODS TO MANAGE PROJECTS

2018 ◽  
pp. 253-257
Author(s):  
Tetiana Yakovenko ◽  
Anastasiia Pustovit
Keyword(s):  
Selection Criteria ◽  
Critical Parameters ◽  
Optimal Selection ◽  
Mathematical Methods ◽  
Work Factors ◽  
Proposed Model ◽  
Production Capacities ◽  
The Cost ◽  
Future Work ◽  
Selection Of

Introduction. The deadline for the planned works in the project is one of the critical parameters that are not mostly approached. In addition, non-compliance with the deadline usually leads to an increase in the cost of the project. It can be the failure in fulfilling another critical parameter of the projects. One of the reasons of this problem is the non-optimal appointment of job executors. Purpose. The article aims to develop an economics and mathematical model for optimal selection of project executors under uncertainty. Results. In order to achieve the goal of the work, factors, which influence its scheduled completion, existing models of executors’ optimal selection are analysed. The unsolved parts of the problem are identified. Executors’ selection criteria and significant limitations of the model are analysed. It has been concluded that one of the most critical factors is the uncertainty of the executors’ production capacities at the time of their carrying out (the number of free technics or workers who will be free and may be involved in future work). The proposed model allows choosing the optimal executors not only under the term’s criteria, but also under taking into account the total cost of the planned works. It can also be used by enterprises and organizations that involve a subcontractor to perform certain work.

scholarly journals An Optimized Mathematical Model for Items Supplies Planning of a Logistic System

2016 ◽  
Vol 10 (10) ◽  
pp. 133
Author(s):  
Mohammad Ali Nasiri Khalili ◽  
Mostafa Kafaei Razavi ◽  
Morteza Kafaee Razavi
Keyword(s):  
Mathematical Model ◽  
Mixed Integer ◽  
Logistics System ◽  
Proposed Model ◽  
Multiple Objective Decision Making ◽  
System Requirements ◽  
Purchasing Logistics ◽  
The Cost ◽  
The Mathematical Model ◽  
First Time

Items supplies planning of a logistic system is one of the major issue in operations research. In this article the aim is to determine how much of each item per month from each supplier logistics system requirements must be provided. To do this, a novel multi objective mixed integer programming mathematical model is offered for the first time. Since in logistics system, delivery on time is very important, the first objective is minimization of time in delivery on time costs (including lack and maintenance costs) and the cost of purchasing logistics system. The second objective function is minimization of the transportation supplier costs. Solving the mathematical model shows how to use the Multiple Objective Decision Making (MODM) can provide the ensuring policy and transportation logistics needed items. This model is solved with CPLEX and computational results show the effectiveness of the proposed model.

Progress and Prospects for the Development of Computer-Generated Actors for Military Simulation: Part 1-Introduction and Background

2003 ◽  
Vol 12 (3) ◽  
pp. 311-325 ◽  
Author(s):  
Martin R. Stytz ◽  
Sheila B. Banks
Keyword(s):  
Research Area ◽  
Background Information ◽  
Life Cycle Costs ◽  
Material Management ◽  
Reasoning System ◽  
Force Structure ◽  
The Cost ◽  
Future Work ◽  
Level Training ◽  
Made In

The development of computer-generated synthetic environments, also calleddistributed virtual environments, for military simulation relies heavily upon computer-generated actors (CGAs) to provide accurate behaviors at reasonable cost so that the synthetic environments are useful, affordable, complex, and realistic. Unfortunately, the pace of synthetic environment development and the level of desired CGA performance continue to rise at a much faster rate than CGA capability improvements. This insatiable demand for realism in CGAs for synthetic environments arises from the growing understanding of the significant role that modeling and simulation can play in a variety of venues. These uses include training, analysis, procurement decisions, mission rehearsal, doctrine development, force-level and task-level training, information assurance, cyberwarfare, force structure analysis, sustainability analysis, life cycle costs analysis, material management, infrastructure analysis, and many others. In these and other uses of military synthetic environments, computer-generated actors play a central role because they have the potential to increase the realism of the environment while also reducing the cost of operating the environment. The progress made in addressing the technical challenges that must be overcome to realize effective and realistic CGAs for military simulation environments and the technical areas that should be the focus of future work are the subject of this series of papers, which survey the technologies and progress made in the construction and use of CGAs. In this, the first installment in the series of three papers, we introduce the topic of computer-generated actors and issues related to their performance and fidelity and other background information for this research area as related to military simulation. We also discuss CGA reasoning system techniques and architectures.

Mathematical Modeling and Analysis of Covid-19 Infection spreads in India with Restricted Optimal Treatment on Disease Incidence

BIOMATH ◽  
2021 ◽  
Vol 10 (1) ◽  
pp. 2106147
Author(s):  
Debkumar Pal ◽  
D Ghosh ◽  
P K Santra ◽  
G S Mahapatra
Keyword(s):  
Mathematical Modeling ◽  
Objective Function ◽  
Reproduction Number ◽  
Disease Incidence ◽  
Treatment Control ◽  
Modeling And Analysis ◽  
Proposed Model ◽  
The Cost ◽  
Disease Free ◽  
The Basic Reproduction Number

This paper presents the current situation and how to minimize its effect in India through a mathematical model of infectious Coronavirus disease (COVID-19). This model consists of six compartments to population classes consisting of susceptible, exposed, home quarantined, government quarantined, infected individuals in treatment, and recovered class. The basic reproduction number is calculated, and the stabilities of the proposed model at the disease-free equilibrium and endemic equilibrium are observed. The next crucial treatment control of the Covid-19 epidemic model is presented in India's situation. An objective function is considered by incorporating the optimal infected individuals and the cost of necessary treatment. Finally, optimal control is achieved that minimizes our anticipated objective function. Numerical observations are presented utilizing MATLAB software to demonstrate the consistency of present-day representation from a realistic standpoint.

scholarly journals Valuation of Guaranteed Maximum Price Contracts in Korea Using the Collar Option Model

2018 ◽  
Vol 2018 ◽  
pp. 1-9
Author(s):  
Sanghyo Lee ◽  
Hyeongjae Jang ◽  
Yonghan Ahn
Keyword(s):  
Profit Sharing ◽  
Construction Companies ◽  
Construction Market ◽  
Positive Effects ◽  
Valuation Model ◽  
Maximum Price ◽  
Proposed Model ◽  
Option Model ◽  
Payoff Structure ◽  
Future Work

This study assessed the levels of risk that contractors may be subject to while executing a GMP contract by applying a collar option model to the case study of an apartment project in Korea and identified implications for the application of GMP contracts in Korea. The payoff structure of the GMP contract was defined based on the collar option model and a profit sharing ratio calculated to evaluate the risks involved in GMP contracts. The results showed that an increase in the GMP and a decrease in the expected cost and cost range were accompanied by a decrease in the profit sharing ratio. The proposed valuation model for GMP contracts is expected to help clients and contractors in Korea negotiate reasonable contracts as it enables the contractor to utilize the proposed model as basic data, the client to evaluate the performance of the contractor, and both parties to agree a reasonable profit sharing ratio. Implementing GMP contracts with CMR is likely to have a number of positive effects on the Korean construction market. However, in order to maximize these effects, it is necessary to have the ability to evaluate cost uncertainty. Accordingly, it is very important to analyze the factors that influence cost volatility. In future work, the various factors that have an impact on the GMP must be studied to maximize the positive effects of the framework proposed in this paper. An analysis of the effect of each factor on the change in the GMP will help Korean construction companies who are attempting to introduce GMP contracts to perform their preconstruction services effectively.

scholarly journals A Scorecard to Detect Financial Leverage Profitability

2018 ◽  
Vol 13 (3) ◽  
pp. 244
Author(s):  
Laura Broccardo ◽  
Luisa Tibiletti ◽  
Pertti Vilpas
Keyword(s):  
Net Present Value ◽  
Economic Value ◽  
Weighted Average ◽  
Cost Of Debt ◽  
Corporate Management ◽  
External Financing ◽  
Present Value ◽  
Financial Decisions ◽  
Proposed Model ◽  
The Cost

This study investigates how balancing internal and external financing sources can create economic value. We set a financial scorecard, consisting of the Cost of Debt (COD), Return on Investment (ROI), and the Cost of Equity (COE). We show that COE should be a cap for COD and a floor for ROI in order to increase the Net Present Value at Weighted Average Cost of Capital and the Adjusted Present Value of the levered investment. However, leverage should be carefully monitored if COD and ROI go off the grid. Situations where leverage has the opposite effect on value creation and the Equity Internal Rate of Return are also discussed. Illustrative examples are given. The proposed model aims to help corporate management in financial decisions.

The Finnish Climate Change Act: Architecture, Functions, and Challenges

Climate Law ◽  
2014 ◽  
Vol 4 (3-4) ◽  
pp. 301-326 ◽  
Author(s):  
Ismo Pölönen
Keyword(s):  
Climate Change ◽  
Climate Policy ◽  
Policy Making ◽  
Dynamic Features ◽  
Policy Choices ◽  
Legal Implications ◽  
Proposed Model ◽  
Term Energy ◽  
Public Scrutiny ◽  
The Cost

The article examines the key features and functions of the proposed Finnish Climate Change Act (fcca). It also analyses the legal implications of the Act and the qualities and factors which may limit its effectiveness. The paper argues that, despite its weak legal implications, the fcca would provide the regulatory preconditions for higher-quality climate policy-making in Finland, and it has the capacity to play an important role in national climate policy. The fcca would deliver regulatory foundations for systematic and integrated climate policy-making, also enabling wide public scrutiny. The proposed model leaves room for manifold climate-policy choices in varying societal and economical contexts. The cost of dynamic features is the relalow predictability in terms of sectorial paths on emission reductions. Another relevant challenge relates to the intended preparation of overlapping mid-term energy and climate plans with instruments of the fcca.

scholarly journals Evaluation of a Systems-Based Tobacco Cessation Program Using Bedside Volunteers

2018 ◽  
Vol 22 (3) ◽  
pp. 440-445 ◽  
Author(s):  
Denise S Taylor ◽  
Dominique Medaglio ◽  
Claudine T Jurkovitz ◽  
Freda Patterson ◽  
Zugui Zhang ◽  
...  
Keyword(s):  
Tobacco Cessation ◽  
Patient Characteristics ◽  
Post Discharge ◽  
Cessation Treatment ◽  
Opt Out ◽  
Artery Disease ◽  
Intent To Treat ◽  
The Cost ◽  
Referral Completion ◽  
Future Work

Abstract Introduction Hospitalization and post-discharge provide an opportune time for tobacco cessation. This study tested the feasibility, uptake, and cessation outcomes of a hospital-based tobacco cessation program, delivered by volunteers to the bedside with post-discharge referral to Quitline services. Patient characteristics associated with Quitline uptake and cessation were assessed. Methods Between February and November 2016, trained hospital volunteers approached inpatient tobacco users on six pilot units. Volunteers shared a cessation brochure and used the ASK-ADVISE-CONNECT model to connect ready to quit patients to the Delaware Quitline via fax-referral. Volunteers administered a follow-up survey to all admitted tobacco users via telephone or email at 3-months post-discharge. Results Of the 743 admitted tobacco users, 531 (72%) were visited by a volunteer, and 97% (531/547) of those approached, accepted the visit. Over one-third (201/531; 38%) were ready to quit and fax-referred to the Quitline, and 36% of those referred accepted Quitline services. At 3 months post-discharge, 37% (135/368) reported not using tobacco in the last 30 days; intent-to-treat cessation rate was 18% (135/743). In a multivariable regression model of Quitline fax-referral completion, receiving nicotine replacement therapy (NRT) during hospitalization was the strongest predictor (odds ratios [OR] = 1.97; 95% confidence interval [CI] = 1.34 to 2.90). In a model of 3-month cessation, receiving Quitline services (OR = 3.21, 95% CI = 1.35 to 7.68) and having coronary artery disease (OR = 2.28; 95% CI = 1.11 to 4.68) were associated with tobacco cessation, but a volunteer visit was not. Conclusions An “opt-out” tobacco cessation service using trained volunteers is feasible for connecting patients to Quitline services. Implications This study demonstrates the feasibility of a systems-based approach to link inpatients to evidence-based treatment for tobacco use. This model used trained bedside volunteers to connect inpatients to a state-funded Quitline after discharge that offers free cessation treatment of telephone coaching and cessation medications. Receiving NRT during hospitalization positively impacted Quitline referral, and engagement with Quitline resources was critical to tobacco abstinence post-discharge. Future work is needed to evaluate the cost-effectiveness and sustainability of this volunteer model.

scholarly journals Self-Supervised Mixture-of-Experts by Uncertainty Estimation

2019 ◽  
Vol 33 ◽  
pp. 5933-5940
Author(s):  
Zhuobin Zheng ◽  
Chun Yuan ◽  
Xinrui Zhu ◽  
Zhihui Lin ◽  
Yangyang Cheng ◽  
...  
Keyword(s):  
Common Knowledge ◽  
Uncertainty Estimation ◽  
Complex Environments ◽  
Mixture Of Experts ◽  
Predictive Uncertainty ◽  
Value Evaluation ◽  
Policy Gradient ◽  
Experience Replay ◽  
Gating Network ◽  
Efficient Transfer

Learning related tasks in various domains and transferring exploited knowledge to new situations is a significant challenge in Reinforcement Learning (RL). However, most RL algorithms are data inefficient and fail to generalize in complex environments, limiting their adaptability and applicability in multi-task scenarios. In this paper, we propose SelfSupervised Mixture-of-Experts (SUM), an effective algorithm driven by predictive uncertainty estimation for multitask RL. SUM utilizes a multi-head agent with shared parameters as experts to learn a series of related tasks simultaneously by Deep Deterministic Policy Gradient (DDPG). Each expert is extended by predictive uncertainty estimation on known and unknown states to enhance the Q-value evaluation capacity against overfitting and the overall generalization ability. These enable the agent to capture and diffuse the common knowledge across different tasks improving sample efficiency in each task and the effectiveness of expert scheduling across multiple tasks. Instead of task-specific design as common MoEs, a self-supervised gating network is adopted to determine a potential expert to handle each interaction from unseen environments and calibrated completely by the uncertainty feedback from the experts without explicit supervision. To alleviate the imbalanced expert utilization as the crux of MoE, optimization is accomplished via decayedmasked experience replay, which encourages both diversification and specialization of experts during different periods. We demonstrate that our approach learns faster and achieves better performance by efficient transfer and robust generalization, outperforming several related methods on extended OpenAI Gym’s MuJoCo multi-task environments.

Sign in / Sign up

Export Citation Format

Share Document