Application of Meta-Heuristic Algorithms for Training Neural Networks and Deep Learning Architectures: A Comprehensive Review

Neural Processing Letters, Journal Year: 2022, Volume and Issue: 55(4), P. 4519 - 4622

Published: Oct. 31, 2022

Language: Английский

---

Tuning Machine Learning Models Using a Group Search Firefly Algorithm for Credit Card Fraud Detection

Mathematics, Journal Year: 2022, Volume and Issue: 10(13), P. 2272 - 2272

Published: June 29, 2022

Recent advances in online payment technologies combined with the impact of COVID-19 global pandemic has led to a significant escalation number transactions and credit card payments being executed every day. Naturally, there also been an frauds, which is having on banking institutions, corporations that issue cards, finally, vendors merchants. Consequently, urgent need implement establish proper mechanisms can secure integrity transactions. The research presented this paper proposes hybrid machine learning swarm metaheuristic approach address challenge fraud detection. novel, enhanced firefly algorithm, named group search was devised then used tune support vector machine, extreme gradient-boosting models. Boosted models were tested real-world detection dataset, gathered from European users. original dataset highly imbalanced; further analyze performance tuned models, second experiment performed for purpose research, expanded by utilizing synthetic minority over-sampling approach. proposed compared other recent state-of-the-art approaches. Standard indicators have evaluation, such as accuracy classifier, recall, precision, area under curve. experimental findings clearly demonstrate algorithm obtained superior results comparison hybridized competitor metaheuristics.

Language: Английский

---

Novel hybrid firefly algorithm: an application to enhance XGBoost tuning for intrusion detection classification

PeerJ Computer Science, Journal Year: 2022, Volume and Issue: 8, P. e956 - e956

Published: April 29, 2022

The research proposed in this article presents a novel improved version of the widely adopted firefly algorithm and its application for tuning optimising XGBoost classifier hyper-parameters network intrusion detection. One greatest issues domain detection systems are relatively high false positives negatives rates. In study, by using optimised with algorithm, challenge is addressed. Based on established practice from modern literature, was first validated 28 well-known CEC2013 benchmark instances comparative analysis original other state-of-the-art metaheuristics conducted. Afterwards, devised method tested optimisation tuned used benchmarking NSL-KDD dataset more recent USNW-NB15 Obtained experimental results prove that has significant potential tackling machine learning it can be improving classification accuracy average precision systems.

Language: Английский

---

Novel Improved Salp Swarm Algorithm: An Application for Feature Selection

Sensors, Journal Year: 2022, Volume and Issue: 22(5), P. 1711 - 1711

Published: Feb. 22, 2022

We live in a period when smart devices gather large amount of data from variety sensors and it is often the case that decisions are taken based on them more or less autonomous manner. Still, many inputs do not prove to be essential decision-making process; hence, utmost importance find means eliminating noise concentrating most influential attributes. In this sense, we put forward method swarm intelligence paradigm for extracting important features several datasets. The thematic paper novel implementation an algorithm branch machine learning domain improving feature selection. combination with metaheuristic approaches has recently created new artificial called learnheuristics. This approach benefits both capability selection solutions impact accuracy performance, as well known characteristic algorithms efficiently comb through search space solutions. latter used wrapper improvements significant. paper, modified version salp proposed. solution verified by 21 datasets classification model K-nearest neighborhoods. Furthermore, performance compared best same test setup resulting better number proposed solution. Therefore, tackles demonstrates its success benchmark

Language: Английский

---

Metaheuristic-Based Hyperparameter Tuning for Recurrent Deep Learning: Application to the Prediction of Solar Energy Generation

Axioms, Journal Year: 2023, Volume and Issue: 12(3), P. 266 - 266

Published: March 4, 2023

As solar energy generation has become more and important for the economies of numerous countries in last couple decades, it is highly to build accurate models forecasting amount green that will be produced. Numerous recurrent deep learning approaches, mainly based on long short-term memory (LSTM), are proposed dealing with such problems, but most may differ from one test case another respect architecture hyperparameters. In current study, use an LSTM a bidirectional (BiLSTM) data collection that, besides time series values denoting generation, also comprises corresponding information about weather. The research additionally endows hyperparameter tuning by means enhanced version recently metaheuristic, reptile search algorithm (RSA). output tuned neural network compared ones several other state-of-the-art metaheuristic optimization approaches applied same task, using experimental setup, obtained results indicate approach as better alternative. Moreover, best model achieved R2 0.604, normalized MSE value 0.014, which yields improvement around 13% over traditional machine models.

Language: Английский

---

Optical microscope algorithm: A new metaheuristic inspired by microscope magnification for solving engineering optimization problems

Knowledge-Based Systems, Journal Year: 2023, Volume and Issue: 279, P. 110939 - 110939

Published: Sept. 3, 2023

Language: Английский

---

Decomposition aided attention-based recurrent neural networks for multistep ahead time-series forecasting of renewable power generation

PeerJ Computer Science, Journal Year: 2024, Volume and Issue: 10, P. e1795 - e1795

Published: Jan. 18, 2024

Renewable energy plays an increasingly important role in our future. As fossil fuels become more difficult to extract and effectively process, renewables offer a solution the ever-increasing demands of world. However, shift toward renewable is not without challenges. While reliable means storage that can be converted into usable energy, are dependent on external factors used for generation. Efficient often relying batteries have limited number charge cycles. A robust efficient system forecasting power generation from sources help alleviate some difficulties associated with transition energy. Therefore, this study proposes attention-based recurrent neural network approach generated sources. To networks make accurate forecasts, decomposition techniques utilized applied time series, modified metaheuristic introduced optimized hyperparameter values networks. This has been tested two real-world datasets covering both solar wind farms. The models by metaheuristics were compared those produced other state-of-the-art optimizers terms standard regression metrics statistical analysis. Finally, best-performing model was interpreted using SHapley Additive exPlanations.

Language: Английский

---

The explainable potential of coupling hybridized metaheuristics, XGBoost, and SHAP in revealing toluene behavior in the atmosphere

The Science of The Total Environment, Journal Year: 2024, Volume and Issue: 929, P. 172195 - 172195

Published: April 15, 2024

Language: Английский

---

Multi-Swarm Algorithm for Extreme Learning Machine Optimization

Sensors, Journal Year: 2022, Volume and Issue: 22(11), P. 4204 - 4204

Published: May 31, 2022

There are many machine learning approaches available and commonly used today, however, the extreme is appraised as one of fastest and, additionally, relatively efficient models. Its main benefit that it very fast, which makes suitable for integration within products require models taking rapid decisions. Nevertheless, despite their large potential, they have not yet been exploited enough, according to recent literature. Extreme machines still face several challenges need be addressed. The most significant downside performance model heavily depends on allocated weights biases hidden layer. Finding its appropriate values practical tasks represents an NP-hard continuous optimization challenge. Research proposed in this study focuses determining optimal or near layer specific tasks. To address task, a multi-swarm hybrid approach has proposed, based three swarm intelligence meta-heuristics, namely artificial bee colony, firefly algorithm sine-cosine algorithm. method thoroughly validated seven well-known classification benchmark datasets, obtained results compared other already existing similar cutting-edge from simulation point out suggested technique capable obtain better generalization than rest included comparative analysis terms accuracy, precision, recall, f1-score indicators. Moreover, prove combining two algorithms effective joining approaches, additional hybrids generated by pairing, each, methods employed approach, were also implemented against four challenging datasets. findings these experiments superior Sample code devised ELM tuning framework GitHub.

Language: Английский

---

The AdaBoost Approach Tuned by Firefly Metaheuristics for Fraud Detection

2022 IEEE World Conference on Applied Intelligence and Computing (AIC), Journal Year: 2022, Volume and Issue: unknown, P. 834 - 839

Published: June 17, 2022

The use of powerful classifiers is broad and the problem fraud detection tends to benefit from similar solutions as well. in digital age cannot be disregarded number cases worrisome. machine learning has been beneficial many real-world problems, classification ability such high. Furthermore, these are not without shortcomings, possibilities hybrid methods explored for reasons further enhancements. Therefore, research proposed this manuscript, adaptive boosting algorithm optimized by firefly metaheuristics validated against imbalanced credit card dataset. Moreover, synthetic minority over-sampling technique applied addressing class imbalance. According experimental findings, method shows substantially better performance than other state-of-the-art models tackling same terms standard metrics.

Language: Английский

---