Corona Virus Optimization Algorithm (CVOA): A Successful Application in Hybrid Approaches to Find Parameters in Machine Learning Model Optimization

1. Introduction

The corona virus (COVID-19) is a new respiratory virus, firstly discovered in humans in December 2019 in Wuhan china that has spread worldwide, having been reported more than 1 million infected people so far [1]. Corona viruses are a large family of viruses which may cause illness in animals or humans [2].  In humans, several corona viruses are known to cause respiratory infections ranging from the common cold to more severe diseases such as Middle East Respiratory Syndrome (MERS) and Severe Acute Respiratory Syndrome (SARS). There are currently no approved human corona virus vaccines [3]. Future considerations in development of a vaccine are on processing.

2. Structure of Corona virus

Corona virus is spherical shaped to pleomorphic enveloped particles [4]. The envelope is studded with projecting glycoproteins, and surrounds a core consisting of matrix protein enclosed within which a single strand of positive-sense RNA is associated with nucleoprotein [5]. The envelope glycoprotein’s are responsible for attachment to the host cell and also carry the main antigenic epitopes, particularly the epitopes recognized by neutralizing antibodies [6].

3. Inspiration of CVOA

The main inspiration of CVOA is oriented with how the corona virus spreads and infects healthy people [7].

4. A Comprehensive Review about Corona Virus

5. The Global impact of Corona virus in Real world

The global economy from COVID-19 has been both faster and more severe global financial crisis (GFC) and even the Great Depression in the real world situations [8].  Businesses that invest in strategic, operational and financial resilience to emerging global risks will be better positioned to respond and recover [9].

6. Implementation of CVOA

Step 1. Generation of the initial population.

The initial population consists of one individual, the so-called Zero Patient (ZP). As in the corona virus epidemic, it identifies the first human being infected.

Step 2. Disease propagation

Two types of spreading are considered, according to a given probability Ordinary spreaders. Infected individuals will infect new ones according to the corona virus spreading rate

Step 3. Updating populations.

Three populations are maintained and updated for each generation. Dead population, Recovered population and New infected population.

Step 4. Stop criterion.

CVOA ability to end without the need of controlling any parameter. A preset number of iterations can be added to the stop criterion. The social isolation measures also contribute to reaching the stop criterion [10].

7. Hybridizing deep learning with CVOA

The term hybridize is used in this context as the combination of  two computational techniques deep learning and CVOA so that the best hyper parameter values are discovered.Bioinspired models typically mimic behaviors from the nature and are known for their successful application in hybrid approaches to find parameters in machine learning model optimization [11]. Hence, the individual codification has been implemented in order to apply CVOA to optimize deep neural network architectures as shown in the fig below.

8. Pseudo code of CVOA

9. Advantages & Disadvantages of CVOA

10. Application of CVOA

In CVOA the problem of electricity load time series forecasting has been addressed, showing quite remarkable performance [12].

Reference

[1] R. Crespo-Cano, S. Cuenca-Asensi, E. Fernandez and Martínez-Álvarez, “Metaheuristic Optimisation Algorithms for Tuning a Bioinspired Retinal Model”, Sensors, vol. 19, no. 22, p. 4834, 2019. Available: 10.3390/s19224834.

 [2] Glover, J. Hao and G. Kochenberger, “Polynomial unconstrained binary optimisation â part 2”, International Journal of Metaheuristics, vol. 1, no. 4, p. 317, 2011. Available: 10.1504/ijmheur.2011.044356.

[3]J. Bedi and D. Toshniwal, “Deep learning framework to forecast electricity demand”, Applied Energy, vol. 238, pp. 1312-1326, 2019. Available: 10.1016/j.apenergy.2019.01.113.

[4]A. Bosire, “Recurrent Neural Network Training using ABC Algorithm for Traffic Volume Prediction”, Informatica, vol. 43, no. 4, 2019. Available: 10.31449/inf.v43i4.2709.

[5]L. Calvet, J. Armas, D. Masip and A. Juan, “Learnheuristics: hybridizing metaheuristics with machine learning for optimization with dynamic inputs”, Open Mathematics, vol. 15, no. 1, pp. 261-280, 2017. Available: 10.1515/math-2017-0029.

[6]Y. Xu, X. Wei and S. Chen, “Research on Railway Passenger Volume Prediction Based on LSTM Neural Network”, IOP Conference Series: Materials Science and Engineering, vol. 688, p. 044017, 2019. Available: 10.1088/1757-899x/688/4/044017.

[7]H. Chung and K. Shin, “Genetic Algorithm-Optimized Long Short-Term Memory Network for Stock Market Prediction”, Sustainability, vol. 10, no. 10, p. 3765, 2018. Available: 10.3390/su10103765.

[8]A. Darwish, A. Hassanien and S. Das, “A survey of swarm and evolutionary computing approaches for deep learning”, Artificial Intelligence Review, vol. 53, no. 3, pp. 1767-1812, 2019. Available: 10.1007/s10462-019-09719-2.

[9]S. De Cnudde, Y. Ramon, D. Martens and F. Provost, “Deep Learning on Big, Sparse, Behavioral Data”, Big Data, vol. 7, no. 4, pp. 286-307, 2019. Available: 10.1089/big.2019.0095.

[10]A. ElSaid, F. El Jamiy, J. Higgins, B. Wild and T. Desell, “Optimizing long short-term memory recurrent neural networks using ant colony optimization to predict turbine engine vibration”, Applied Soft Computing, vol. 73, pp. 969-991, 2018. Available: 10.1016/j.asoc.2018.09.013.

[11]D. Devikanniga, K. Vetrivel and N. Badrinath, “Review of Meta-Heuristic Optimization based Artificial Neural Networks and its Applications”, Journal of Physics: Conference Series, vol. 1362, p. 012074, 2019. Available: 10.1088/1742-6596/1362/1/012074.

[12]Y. Liang and J. Cuevas Juarez, “A self-adaptive virus optimization algorithm for continuous optimization problems”, Soft Computing, 2020. Available: 10.1007/s00500-020-04730-0.