Machine Learning Methods for Ranking

Rahangdale, Ashwini; Raut, Shital

doi:10.1142/s021819401930001x

Cited by 17 publications

(11 citation statements)

References 70 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For classification problems, it is common to use the Logistic Regression or Maximum-Entropy method. Although, when used for learning-to-rank problems, other methods like Support Vector Machine (SVM) seem to perform better 11 . The Logistic Regression is a well known method for classification, which models its outcomes as probabilities using the logistic function.…”

Section: Information Retrievalmentioning

confidence: 99%

An empirical study of learning-to-rank for spare parts consumption in the repair process

Duarte¹,

Moraes²,

Padula³

2023

Preprint

View full text Add to dashboard Cite

The repair process of devices is an important part of the business of many original equipment manufacturers. The consumption of spare parts, during the repair process, is driven by the defects found during inspection of the devices, and these parts are a big part of the costs in the repair process. In previous work we proposed a data-driven method for Supply Chain Control Tower solutions to provide support for the automatic check of spare parts consumption in the repair process. In this paper, we continue our investigation of a multi-label classification problem and explore alternatives in the learning-to-rank approach, where we simulate the passage of time using more data while training and comparing hundreds of Machine Learning models to provide an automatic check in the consumption of spare parts. We investigate the effects of different train set sizes, retraining intervals, models and hyper-parameter search using Bayesian Optimization. The results show that we were able to improve the trained models and achieve a higher mean NDCG@20 score of 86% when ranking the expected parts. Focusing on the most recent data, we achieve a NDCG@20 score of 90%, while obtaining a ratio of marked parts of just 4% of the consumed parts for use in alert generation.

show abstract

Section: Information Retrievalmentioning

confidence: 99%

An empirical study of learning-to-rank for spare parts consumption in the repair process

Duarte¹,

Moraes²,

Padula³

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…And for classification problems, it is common to use the Logistic Regression or Maximum‐Entropy method. Although, when used for learning‐to‐rank problems, other methods like support vector machine (SVM) seem to perform better (Rahangdale & Raut, 2019).…”

Section: Related Workmentioning

confidence: 99%

“…In classification problems, a single label is assigned to each document but sometimes we need to assign many labels to each document, in this case we can employ the multi‐label classification extension (Tsoumakas & Katakis, 2007). To evaluate the models in multi‐label classification, the most common metrics are Accuracy, Normalized Discounted Cumulative Gain, and Mean Average Precision scores (Järvelin & Kekäläinen, 2002; Rahangdale & Raut, 2019; Tsoumakas & Katakis, 2007). Additionally, we can combine predictions of many methods in two main ways: when working with many instances of the same base method, we can use ensemble methods to improve generalization and robustness, and when working with conceptually different methods, we can use any of several output fusion methods (Sagi & Rokach, 2018).…”

Section: Related Workmentioning

confidence: 99%

An empirical study of learning‐to‐rank for spare parts consumption in the repair process

Duarte,

de Haro Moraes,

Padula

2023

Expert Systems

View full text Add to dashboard Cite

The repair process of devices is an important part of the business of many original equipment manufacturers. The consumption of spare parts, during the repair process, is driven by the defects found during inspection of the devices, and these parts are a big part of the costs in the repair process. In previous work, we proposed a data‐driven method for Supply Chain Control Tower solutions to provide support for the automatic check of spare parts consumption in the repair process. In this article, we continue our investigation of a multi‐label classification problem and explore alternatives in the learning‐to‐rank approach, where we simulate the passage of time using more data while training and comparing hundreds of Machine Learning models to provide an automatic check in the consumption of spare parts. We investigate the effects of different train set sizes, retraining intervals, models and hyper‐parameter search using Bayesian Optimization. We define a custom metric, the Ratio of Marked Parts, measuring how many spare parts are marked for revision at the end of the repair process. The results show that we were able to improve the trained models and achieve a higher mean NDCG@20 score of 86% when ranking the expected parts. While focusing on the most recent data, we achieved a NDCG@20 score of 90% and obtained a Ratio of Marked Parts of just 4% of the consumed parts for use in alert generation.

show abstract

“…The whole image is then analyzed through machine learning outside of the smartphone. Machine learning frameworks such as Learning-to-rank can be used to analyze the resulting image [6]. An enzyme-free nucleic acid amplification method for detecting influenza virus DNA and miR-316 microRNA was developed in 2018 [7].…”

Section: Introductionmentioning

confidence: 99%

MoFlus: An Open-Source Android Software for Fluorescence-Based Point of Care

Wirawan

Wibowo

2022

JBIOMES

View full text Add to dashboard Cite

High-sensitivity fluorescence-based tests are utilized to monitor various activities in life science research. These tests are specifically used as health monitoring tools to detect diseases. Fluorescence-based test facilities in rural areas and developing countries, however, remain limited. Point-of-care (POC) tests based on fluorescence detection have become a solution to the limitations of fluorescence-based tools in developing countries. POC software for smartphone cameras was generally developed for specific devices and tools, and it ability to select the desired region of interest (ROI) is limited. In this work, we developed Mobile Fluorescence Spectroscopy (MoFlus), an open-source Android software for camera-based POC. We mainly aimed to develop camera-based POC software that can be used for the dynamic selection of ROI; the number of samples; and the types of detection, color, data, and for communication with servers. MoFlus facilitated the use of touch screens and data given that it was developed on the basis of the SurfaceView library in Android and Javascript object notation applications. Moreover, the function and endurance of the app when used multiple times and with different numbers of images were tested.

show abstract

Machine Learning Methods for Ranking

Cited by 17 publications

References 70 publications

An empirical study of learning-to-rank for spare parts consumption in the repair process

An empirical study of learning-to-rank for spare parts consumption in the repair process

An empirical study of learning‐to‐rank for spare parts consumption in the repair process

MoFlus: An Open-Source Android Software for Fluorescence-Based Point of Care

Contact Info

Product

Resources

About