Person:
ALBEY, Erinç

Profile Picture

Email Address

Birth Date

WoSScopusGoogle ScholarORCID

Name

Job Title

First Name

Erinç

Last Name

ALBEY
OrgUnit Logo
Organizational Unit

Filters

2016 - 2019142020 - 202410
Reset filters

Settings

Publication Search Results

Now showing 1 - 10 of 24
  • Placeholder
    ArticlePublicationMetadata only
    Optimal pricing and inventory strategies for fashion products under time-dependent interest rate and demand
    (Elsevier, 2021-04) Akan, M.; Albey, Erinç; Güler, M. G.; Industrial Engineering; ALBEY, Erinç
    In this work we consider the dynamic pricing problem of a retailer operating in a market with a single fashion item and under time-dependent interest rate. The demand is assumed to be deterministic and dependent on the price and decay with time, i.e., the market shrinks throughout the horizon. Using an optimal-control-theoretic approach, we analytically derive the optimal pricing and inventory strategy for the retailer over a finite horizon setting. We further analyze the ramifications of the optimal pricing decision for different initial inventory levels dictated by the relationship between the supplier and the retailer; and for varying market interest rates. Optimal dynamic pricing policy is a continuous function, which is almost impossible to use in practice. This is handled using approximate piece-wise constant pricing policies. The trade-off between dynamic pricing policy and approximate policies is also investigated.
  • Placeholder
    Conference paperPublicationMetadata only
    Ensemble Learning based on Regressor Chains: A Case on Quality Prediction
    (SciTePress, 2019) Demirel, Kenan Cem; Şahin, Ahmet; Albey, Erinç; Industrial Engineering; ALBEY, Erinç; Demirel, Kenan Cem; Şahin, Ahmet
    In this study we construct a prediction model, which utilizes the production process parameters acquired from a textile machine and predicts the quality characteristics of the final yarn. Several machine learning algorithms (decision tree, multivariate adaptive regression splines and random forest) are used for prediction. An ensemble method, using the idea of regressor chains, is developed to further improve the prediction performance. Collected data is first segmented into two parts (labeled as “normal” and “unusual”) using local outlier factor method, and performance of the algorithms are tested for each segment separately. It is seen that ensemble idea proves its competence especially for the cases where the collected data is categorized as unusual. In such cases ensemble algorithm improves the prediction accuracy significantly. Copyright © 2019 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
  • Thumbnail Image
    Conference paperPublicationOpen Access
    A hierarchical approach for solving simultaneous lot sizing and scheduling problem with secondary resources
    (Elsevier, 2019) Şafak, C. U.; Yılmaz, Görkem; Albey, Erinç; Industrial Engineering; ALBEY, Erinç; YILMAZ, Görkem
    This study represents a decomposition heuristic approach for simultaneous lot sizing and scheduling problem for multiple product, multiple parallel machines with secondary resources. The motivation of the study comes from the real-world instance of a plastic injection plant at Vestel Electronics. The plastic injection plant requires plastic injection molds at the planner's disposal, in order to produce variations of products, by the compatible plastic injection machines. The variations on the molds and the mold changes on the machines bring out sequence dependent major and minor setups. Since each machine requires an operator, we have extended the formulation with workforce and shift planning Results show that proposed heuristic yields comparable solutions to that of exact model for small and medium size instances; and provides schedules for the large size instances, for which exact model cannot find a feasible solution in the allotted time.
  • Placeholder
    Conference paperPublicationMetadata only
    A mathematical model for customer lifetime value based offer management
    (Springer, 2018) Şahin, Ahmet; Can, Zehra; Albey, Erinç; Industrial Engineering; Filipe, J.; Quix, C.; Bernardino, J.; ALBEY, Erinç; Şahin, Ahmet; Can, Zehra
    Customers with prepaid lines possess higher attrition risk compared to postpaid customers, since prepaid customers do not sign long-term obligatory contracts and may churn anytime. For this reason, mobile operators have to offer engaging benefits to keep prepaid subscribers with the company. Since all such offers incur additional cost, mobile operators face an optimization problem while selecting the most suitable offers for customers at risk. In this study, an offer management framework targeting prepaid customers of a telecommunication company is developed. Proposed framework chooses the most suitable offer for each customer through a mathematical model, which utilizes customer lifetime value and churn risk. Lifetime values are estimated using logistic regression and Pareto/NBD models, and several variants of these models are used to predict churn risks using a large number of customer specific features.
  • Thumbnail Image
    Conference paperPublicationOpen Access
    A Markovian approach for time series prediction for quality control
    (Elsevier, 2019) Şahin, Ahmet; Sayımlar, Ayşe Dilara; Teksan, Zehra Melis; Albey, Erinç; Industrial Engineering; TEKSAN, Zehra Melis; ALBEY, Erinç; Şahin, Ahmet; Sayımlar, Ayşe Dilara
    In this work we aim to predict quality levels of incoming batches of a selected product type to a white goods manufacturer from a third party supplier. We apply a Markov Model that captures the quality level of the incoming batch in order to predict the quality status of the future arrivals. The ultimate aim is to generate reliable predictions for the future incoming batches, so that the manufacturing company could warn its supplier if the predictions indicate a significant deterioration in the quality. Applied methodology is compared to several benchmark approaches and its superior performance is shown using a benchmark dataset from the literature and the dataset provided by the manufacturing company. Proposed algorithm performs better compared to benchmarks in detecting the instances with quality level falling outside the tolerances in the validation data; and proves itself as a promising approach for the company.
  • Placeholder
    Conference paperPublicationMetadata only
    Load dependent lead time modelling: a robust optimization approach
    (IEEE, 2018-01-04) Albey, Erinç; Yanıkoğlu, İhsan; Uzsoy, R.; Industrial Engineering; ALBEY, Erinç; YANIKOĞLU, Ihsan
    Although production planning models using nonlinear CFs have shown promising results for semiconductor wafer fabrication facilities, the lack of an effective methodology for estimating the CFs is a significant obstacle to their implementation. Current practice focuses on developing point estimates using least-squares regression approaches. This paper compares the performance of a production planning model using a multi-dimensional CF and its robust counterpart under several experimental settings. As expected, as the level of uncertainty is increased, the resulting production plan deviates from the optimal solution of the deterministic model. On the other hand, production plans found using the robust counterpart are less vulnerable to parameter estimation errors.
  • Thumbnail Image
    Conference paperPublicationOpen Access
    Effective training methods for automatic musical genre classification
    (SciTePress, 2019) Atsız, Eren; Albey, Erinç; Kayış, Enis; Industrial Engineering; Hammoudi, S.; Quix, C.; Bernardino, J.; ALBEY, Erinç; KAYIŞ, Enis
    Musical genres are labels created by human and based on mutual characteristics of songs, which are also called musical features. These features are key indicators for the content of the music. Rather than predictions by human decisions, developing an automatic solution for genre classification has been a significant issue over the last decade. In order to have automatic classification for songs, different approaches have been indicated by studying various datasets and part of songs. In this paper, we suggest an alternative genre classification method based on which part of songs have to be used to have a better accuracy level. Wide range of acoustic features are obtained at the end of the analysis and discussed whether using full versions or pieces of songs is better. Both alternatives are implemented and results are compared. The best accuracy level is 55% while considering the full version of songs. Besides, additional analysis for Turkish songs is also performed. All analysis, data, and results are visualized by a dynamic dashboard system, which is created specifically for the study.
  • Placeholder
    Conference paperPublicationMetadata only
    Churn prediction for mobile prepaid subscribers
    (Institute for Systems and Technologies of Information, Control and Communicatio, 2017) Can, Zehra; Albey, Erinç; Industrial Engineering; ALBEY, Erinç; Can, Zehra
    In telecommunication, mobile operators prefer to acquire postpaid subscribers and increase their incoming revenue based on the usage of postpaid lines. However, subscribers tend to buy and use prepaid mobile lines because of the simplicity of the usage, and due to higher control over the cost of the line compared to postpaid lines. Moreover the prepaid lines have less paper work between the operator and subscriber. The mobile subscriber can end their contract, whenever they want, without making any contact with the operator. After reaching the end of the defined period, the subscriber will disappear, which is defined as “involuntary churn”. In this work, prepaid subscribers’ behavior are defined with their RFM data and some additional features, such as usage, call center and refill transactions. We model the churn behavior using Pareto/NBD model and with two benchmark models: a logistic regression model based on RFM data, and a logistic regression model based on the additional features. Pareto/NBD model is a crucial step in calculating customer lifetime value (CLV) and aliveness of the customers. If Pareto/NBD model proves to be a valid approach, then a mobile operator can define valuable prepaid subscribers using this and decide on the actions for these customers, such as suggesting customized offers.
  • Placeholder
    ArticlePublicationMetadata only
    Neural network estimators for optimal tour lengths of traveling salesperson problem instances with arbitrary node distributions
    (Informs, 2024) Varol, Taha; Özener, Okan Örsan; Albey, Erinç; Industrial Engineering; ÖZENER, Okan Örsan; ALBEY, Erinç; Varol, Taha
    It is essential to solve complex routing problems to achieve operational efficiency in logistics. However, because of their complexity, these problems are often tackled sequentially using cluster-first, route-second frameworks. Unfortunately, such two-phase frameworks can suffer from suboptimality due to the initial phase. To address this issue, we propose leveraging information about the optimal tour lengths of potential clusters as a preliminary step, transforming the two-phase approach into a less myopic solution framework. We introduce quick and highly accurate Traveling Salesperson Problem (TSP) tour length estimators based on neural networks (NNs) to facilitate this. Our approach combines the power of NNs and theoretical knowledge in the routing domain, utilizing a novel feature set that includes node-level, instance-level, and solution-level features. This hybridization of data and domain knowledge allows us to achieve predictions with an average deviation of less than 0.7% from optimality. Unlike previous studies, we design and employ new instances replicating real-life logistics networks and morphologies. These instances possess characteristics that introduce significant computational costs, making them more challenging. To address these challenges, we develop a novel and efficient method for obtaining lower bounds and partial solutions to the TSP, which are subsequently utilized as solution-level predictors. Our computational study demonstrates a prediction error up to six times lower than the best machine learning (ML) methods on their training instances and up to 100 times lower prediction error on out-of-distribution test instances. Furthermore, we integrate our proposed ML models with metaheuristics to create an enumeration-like solution framework, enabling the improved solution of massive scale routing problems. In terms of solution time and quality, our approach significantly outperforms the state-of-the-art solver, demonstrating the potential of our features, models, and the proposed method.
  • Placeholder
    ArticlePublicationMetadata only
    A data-driven matching algorithm for ride pooling problem
    (Elsevier, 2022-04) Şahin, Ahmet; Sevim, İ.; Albey, Erinç; Güler, M. G.; Industrial Engineering; ALBEY, Erinç
    This paper proposes a data-driven matching algorithm for the problem of ride pooling, which is a transportation mode enabling people to share a vehicle for a trip. The problem is considered as a variant of matching problem, since it aims to find a matching between drivers and riders. Proposed algorithm is a machine learning algorithm based on rank aggregation idea, where every feature in a multi-feature dataset provides a ranking of candidate drivers and weight for each feature is learned from past data through an optimization model. Once weight learning and candidate ranking problems are considered simultaneously, resulting optimization model becomes a nonlinear bilevel optimization model, which is reformulated as a single level mixed-integer nonlinear optimization model. To demonstrate the performance of the proposed algorithm, a real-life dataset from a mobile application of a ride pooling start-up company is used and company's current approach is considered as benchmark. Results reveal that proposed algorithm correctly predicts the first choice of riders 17% to 28% better compared to the benchmark in different scenarios. Similarly, proposed algorithm offers recommendation lists in which the preferred driver is ranked 0.38 to 1.12 person closer (to the rider's actual choice) compared to the benchmark.