Abstract: Feature Selection (FS) is an important data management technique that aims to minimize redundant information in a dataset. 
This work proposes DENGO, an improved version of the Northern Goshawk Optimization (NGO), to address the FS problem. The NGO is an efcient swarm-based algorithm that takes its inspiration from the predatory actions of the northern 
goshawk. In order to overcome the disadvantages that NGO is prone to local optimum trap, slow convergence speed and 
low convergence accuracy, two strategies are introduced in the original NGO to boost the efectiveness of NGO. Firstly, a 
learning strategy is proposed where search members learn by learning from the information gaps of other members of the 
population to enhance the algorithm's global search ability while improving the population diversity. Secondly, a hybrid 
diferential strategy is proposed to improve the capability of the algorithm to escape from the trap of the local optimum by 
perturbing the individuals to improve convergence accuracy and speed. To prove the efectiveness of the suggested DENGO, 
it is measured against eleven advanced algorithms on the CEC2015 and CEC2017 benchmark functions, and the obtained 
results demonstrate that the DENGO has a stronger global exploration capability with higher convergence performance and 
stability. Subsequently, the proposed DENGO is used for FS, and the 29 benchmark datasets from the UCL database prove 
that the DENGO-based FS method equipped with higher classifcation accuracy and stability compared with eight other 
popular FS methods, and therefore, DENGO is considered to be one of the most prospective FS techniques. DENGO's code 
can be obtained at https://www.mathworks.com/matlabcentral/fleexchange/158811-project1.

Key words: Northern goshawk optimization , · Learning strategy , · Hybrid diferential strategy , · Numerical optimization , · Feature selection