Feature selection and feature learning in machine learning applications for gas turbines: A review

Xie, Jiarui; Sage, Manuel; Zhao, Yaoyao Fiona

doi:10.1016/j.engappai.2022.105591

Cited by 38 publications

(6 citation statements)

References 116 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Feature engineering has emerged as a crucial technique in machine learning due to the challenges associated with raw data and the need to extract meaningful information for effective model training and prediction. It refers to the process of transforming raw data into a format that is suitable for machine learning algorithms to effectively learn patterns and make accurate predictions [36][37][38]. It involves creating new features, modifying, or transforming existing ones to enhance the predictive power of machine learning models by providing them with more informative and representative features [39].…”

Section: Feature Engineeringmentioning

confidence: 99%

“…Feature engineering could be considered as a generic term for any tasks used between gathering raw data and starting model training, and the selection of the appropriate approach is dependent on the ML model [34,37,40]. Feature engineering encompasses the entire process of preparing and enhancing the features for machine learning.…”

Section: Various Feature Engineering Techniquesmentioning

confidence: 99%

“…It seeks to choose an appropriate subset of features for the ML model from the available input. Reduced noise, storage needs, and training time enhanced data visualization and comprehension, and enhanced prediction performance are some of the possible advantages of feature selection [34,37]. A feature set is defined as A = { x1, x2, x3, ... xn}, where xi represents the i-th feature in the initial feature space Rn.…”

Section: Various Feature Engineering Techniquesmentioning

confidence: 99%

See 2 more Smart Citations

Generative AI-enabled Knowledge Base Fine-tuning: Enhancing Feature Engineering for Customer Churn

Shahabikargar,

Beheshti,

Mansoor

et al. 2024

Preprint

View full text Add to dashboard Cite

Customers are the most critical component in a business’s success regardless of the industry or product. Companies make significant efforts to acquire and, more importantly, retain their existing customers. Customer churn is a significant challenge for businesses, leading to financial losses. To address this challenge, understanding customer’s cognitive status, behaviors, and early signs of churn is crucial. However, predictive and ML-based analysis, being fed with proper features that are indicative of a customer’s cognitive status or behavior, is extremely helpful in addressing this challenge. Having practical ML-based analysis relies on a well-developed feature engineering process. Previous churn analytical studies mainly applied feature engineering approaches that leveraged demographic, product usage, and revenue features alone, and there is a lack of research on leveraging the information-rich content from interactions between customers and companies. Considering the effectiveness of applying domain knowledge and human expertise in feature engineering, and motivated by our previous work, we propose a Customer Churn-related Knowledge Base (ChurnKB) to enhance the feature engineering process. In the ChurnKB, we leverage textual data mining techniques for extracting churn-related features from texts created by customers, e.g., emails or chat logs with company agents, reviews on the company’s website, and feedback on social media. We use Generative AI (GAI) to enhance and enrich the structure of the ChurnKB regarding features related to customer churn-related cognitive status, feelings, and behaviors. We also leveraged feedback loops and crowdsourcing to enhance and approve the validity of the proposed ChurnKB and apply it to develop a classifier for customer churn problems.

show abstract

Section: Feature Engineeringmentioning

confidence: 99%

Section: Various Feature Engineering Techniquesmentioning

confidence: 99%

Section: Various Feature Engineering Techniquesmentioning

confidence: 99%

See 1 more Smart Citation

Generative AI-enabled Knowledge Base Fine-tuning: Enhancing Feature Engineering for Customer Churn

Shahabikargar,

Beheshti,

Mansoor

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…It is worth noting that the initial generation of electricity prices can be facilitated by employing algorithms such as deep learning. Such algorithms establish a correlation model between historical electricity prices and power dynamics, as evident in the works by Xie et al (2023) and Liang et al (2020). Moreover, a new purchase power variable of SBs can be brought into the model of fuzzy training to obtain the initial electricity price closer to the optimal value, thereby improving the convergence speed of calculation.…”

Section: Objective Function Modelingmentioning

confidence: 99%

Intelligent optimization algorithm-based electricity pricing strategy for smart building clusters

Wang,

Liao,

et al. 2023

Front. Energy Res.

View full text Add to dashboard Cite

With the continuous infusion of renewable energy sources, smart buildings have evolved from single-load characteristics into dual characteristics with both electric energy production and consumption capability. Concurrently, the peak and off-peak periods of electricity consumption are influenced by climatic factors, which leads to complexity and deviation from the time-of-use tariffs set by electricity markets, which consequently result in a loss of revenue from grid-based electricity sales. Thus, adopting an innovative pricing mechanism to offset the revenue deficit in the grid assumes paramount significance. Built upon a dual-layer framework that employs intelligent optimization algorithms, this study proposes a pricing strategy for introducing the retail electricity provider into smart building clusters with peer-to-peer power sharing as the core. First, the independent operation model of intelligent buildings and electric energy sharing model without the participation of retail power suppliers are respectively established. Subsequently, with the aim to minimize alliance costs, a novel energy sharing pricing model involving retail electricity suppliers is developed, and a combination of particle swarm optimization and alternating direction multiplier methods is used for distributed solutions within a representative model. This approach yields optimal energy sharing transaction volumes and pricing while ensuring the confidentiality of each participating entity. Lastly, from the perspectives of the power grid, retail electricity suppliers, and multi-building smart alliances, this study conducts simulation analyses of key parameters that influence the bargaining effectiveness of retail electricity suppliers. These parameters encompass the upper limit of pricing, market supervision coefficient, and discount coefficient associated with the grid-based electricity sales to suppliers. Through these analyses, the study further validates the efficacy of the proposed strategy.

show abstract

“…A study by Xie et al [13] categorized feature selection and feature-learning methods as the primary process to produce compact features; compact features were addressed as the result of the feature-reduction process. The feature-selection method aims to reduce the dimension by selecting the subset of features with highly discriminant information.…”

Section: Compact Feature Representationmentioning

confidence: 99%

Compact-Fusion Feature Framework for Ethnicity Classification

2023

View full text Add to dashboard Cite

In computer vision, ethnicity classification tasks utilize images containing human faces to extract ethnicity labels. Ethnicity is one of the soft biometric feature categories useful in data analysis for commercial, public, and health sectors. Ethnicity classification begins with face detection as a preprocessing process to determine a human’s presence; then, the feature representation is extracted from the isolated facial image to predict the ethnicity class. This study utilized four handcrafted features (multi-local binary pattern (MLBP), histogram of gradient (HOG), color histogram, and speeded-up-robust-features-based (SURF-based)) as the basis for the generation of a compact-fusion feature. The compact-fusion framework involves optimal feature selection, compact feature extraction, and compact-fusion feature representation. The final feature representation was trained and tested with the SVM One Versus All classifier for ethnicity classification. When it was evaluated in two large datasets, UTKFace and Fair Face, the proposed framework achieved accuracy levels of 89.14%, 82.19%, and 73.87%, respectively, for the UTKFace dataset with four or five classes and the Fair Face dataset with four classes. Furthermore, the compact-fusion feature with a small number of features at 4790, constructed based on conventional handcrafted features, achieved competitive results compared with state-of-the-art methods using a deep-learning-based approach.

show abstract

Feature selection and feature learning in machine learning applications for gas turbines: A review

Cited by 38 publications

References 116 publications

Generative AI-enabled Knowledge Base Fine-tuning: Enhancing Feature Engineering for Customer Churn

Generative AI-enabled Knowledge Base Fine-tuning: Enhancing Feature Engineering for Customer Churn

Intelligent optimization algorithm-based electricity pricing strategy for smart building clusters

Compact-Fusion Feature Framework for Ethnicity Classification

Contact Info

Product

Resources

About