key: cord-0877805-gbld82hl authors: Al-qaness, Mohammed A.A.; Saba, Amal I.; Elsheikh, Ammar H.; Elaziz, Mohamed Abd; Ibrahim, Rehab Ali; Lu, Songfeng; Hemedan, Ahmed Abdelmonem; Shanmugan, S.; Ewees, Ahmed A. title: Efficient Artificial Intelligence Forecasting Models for COVID-19 Outbreak in Russia and Brazil date: 2020-11-13 journal: Process Saf Environ Prot DOI: 10.1016/j.psep.2020.11.007 sha: d2be7b34de87c9fae5d025e505ff68dee422b6b4 doc_id: 877805 cord_uid: gbld82hl COVID-19 is a new member of the Coronaviridae family that has serious effects on respiratory, gastrointestinal, and neurological systems. COVID-19 spreads quickly worldwide and affects more than 41.5 million persons (till 23 October 2020). It has a high hazard to the safety and health of people all over the world. COVID-19 has been declared as a global pandemic by the World Health Organization (WHO). Therefore, strict special policies and plans should be made to face this pandemic. Forecasting COVID-19 cases in hotspot regions is a critical issue, as it helps the policymakers to develop their future plans. In this paper, we propose a new short term forecasting model using an enhanced version of the Adaptive Neuro-Fuzzy Inference System (ANFIS). An improved Marine Predators Algorithm (MPA), called Chaotic MPA (CMPA), is applied to enhance the ANFIS and to avoid its shortcomings. More so, we compared the proposed CMPA with three artificial intelligence-based models include the original ANFIS, and two modified versions of ANFIS model using both of the original Marine Predators Algorithm (MPA) and Particle Swarm Optimization (PSO). The forecasting accuracy of the models was compared using different statistical assessment criteria. CMPA significantly outperformed all other investigated models. virus. An autoregressive integrated moving average approach was employed for forecasting the re-67 covered and conrmed COVID-19 cases in Italy [10] . Reasonable Brownian motion, where the predators select these two strategies for optimal foraging. It has been 129 tested using various optimization tasks, and it showed promising performance compared to other 130 optimization approaches. In this study, we apply a chaotic MPA to enhance ANFIS performance and avoid the shortcomings compared the CMPA with existing optimization models to conrm its quality and good per-155 formance. The rest sections of this study are presented as follows. We describe the study area and the 157 collected data in Section 2. Section 3 presents the methods applied in this study. where µ is the generalized Gaussian membership function. where the output of ith nodes from the previous layer is represented by w i . Furthermore, Eq. (5) represents the output of Layer 4: in which f represents the function which combines the parameters and inputs of the networks. r i , q i , 204 and p i represent the consequent parameters of node i. Equation (6) represents the output of Layer 5: The MPA considers predator and prey as search agents because when a predator searches for prey, the prey also searches for food. Thus, at the end of each population (generation), the matrix of the ttest predators (elite matrix) is updated. Eq. (8) formulates the elite and prey (X ) [36]: Thereafter, the prey X position is updated using three stages, called high-velocity ratio, unit In the high-velocity ratio stage or so-called exploration phase, each predator moves faster than 213 X, which is performed at the rst third of the number of iteration (i.e., 1 3 t max ). Thus, the prey can 214 be updated using Eq.(9) and (10): here, R ∈ [0, 1] is a vector of uniform random numbers, and P = 0.5 is a constant number. In this phase, the predator and prey move in the same space, and these movements simulate 220 the processes of searching for prey and food. Therefore, this refers to changing the status of the 221 marine predator algorithm from the exploration phase to exploitation phase. In this stage, both presented in Eqs. (11) and (12), when 1 3 t max < t < 2 3 t max : in which R L contains random numbers which follow the Lévy distributions. Eqs. (11) and (12) CF represents a parameter which controls the step size of movements of the predator, where t max is 231 the total number of iterations (generations). 232 This stage is the latest optimization process that occurs when the predator's movement is faster 234 than the prey's movement. It represents the exploitation phase where t > 2 3 t max , as presented by Eq. (15): Where F AD = 0.2, and U are binary solutions, which can be implemented by creating a random 238 solution, which is converted into binary solutions by using threshold 0.2. More so, r ∈ [0, 1] is a 239 random number, where r 1 and r 2 represent the prey indices. The next process is to generate a set of N solutions, representing the conguration of ANFIS parameters. Each solution is evaluated by constructing the ANFIS network according to its value and applied the training set to the constructed ANFIS. Then predict the output and compute the Root mean square error (RMSE), which is applied as tness value, as in Eq.19: in which T and P are the original target, and the predicted output, respectively. N s refers to the 254 size of the sample. After that, the best conguration/solution is determined, which has the best tness value. Then 256 the X solutions will be updated using the operators of CMPA. This is performed by using the value of 257 R that generated using Eq. In the aforementioned discussion, the predictive capabilities of the proposed models have been 300 evaluated. Now, we will apply these models to forecast the cases of further days ahead. The fore- Estimating instant case fatality rate of covid-19 in china Real-time forecasts and risk assessment of novel coronavirus (covid-381 19) cases: A data-driven analysis Online forecasting of covid-19 cases in nigeria using 383 limited data Development of new hybrid model of 385 discrete wavelet decomposition and autoregressive integrated moving average (arima) models in 386 application to one month forecast the casualties cases of covid-19 Suttearima: Short-term forecasting method, a case: Covid-19 and 389 stock market in spain Prediction of the covid-19 spread in african 391 countries and implications for prevention and controls: a case study in south africa Ai-driven tools for coronavirus outbreak: need of active learning and cross-394 population train/test models on multitudinal/multimodal data Agreeing to disagree: Active learn-397 ing with noisy labels without crowdsourcing A novel method for predicting tensile strength of friction stir welded aa6061 aluminium 401 alloy joints based on hybrid random vector functional link and henry gas solubility optimization, lar still using articial neural network and harris hawks optimizer Noise 407 prediction of axial piston pump based on dierent valve materials using a modied articial 408 neural network model Modeling of friction stir welding 410 process using adaptive neuro-fuzzy inference system integrated with harris hawks optimizer Improved prediction of oscillatory heat transfer 413 coecient for a thermoacoustic heat exchanger using modied adaptive neuro-fuzzy inference 414 system Time series forecasting of covid-19 transmission in canada using 416 lstm networks Forecasting the prevalence of covid-19 outbreak in egypt using 418 nonlinear autoregressive articial neural networks Optimization method for forecasting 421 conrmed cases of covid-19 in china Marine predators algorithm 423 for forecasting conrmed cases of covid-19 in italy, usa, iran and korea A novel hybrid methodology for short-term wind power forecasting 426 based on adaptive neuro-fuzzy inference system Evaluation and forecasting of solar radiation using time se-431 ries adaptive neuro-fuzzy inference system: Seoul city as a case study Prediction of biochar yield using adaptive neuro-fuzzy inference system with particle swarm 435 optimization Oil consumption forecasting using optimized 437 adaptive neuro-fuzzy inference system based on sine cosine algorithm A modied adaptive neuro-fuzzy inference 440 system using multi-verse optimizer algorithm for oil consumption forecasting Improved adaptive neuro-fuzzy inference system using gray wolf 443 optimization: A case study in predicting biochar yield Improving adaptive neuro-fuzzy inference system based 446 on a modied salp swarm algorithm using genetic algorithm to forecast crude oil price Marine predators algorithm: A 449 nature-inspired metaheuristic Modeling of 451 solar energy systems using articial neural network: A comprehensive review All authors declare that they have no conict of interest 349