Table 1 Analysis based on performance.

From: Sparse attention with residual pyramidal depthwise separable convolutional based malware detection with optimization mechanism

Author and year

Technique

Merits

Demerits

Omar et al. 202128

Hybrid model

Reduces the feature space on large scale domain

Detection and classification was complex

Tom et al. 202129

Deep-Hook

Performs trusted detection

Complex method

Jeffrey et al. 2021 30

Recurrent neural networks with LSTM and BIDI

Takes less time and achieves high performance

Samples was low

Donghai et al. 2021 31

Hardware trace

Collects the execution control flow information efficiently

Stealthy malware was not detected

Ce Li et al. 202232

API sequence

Good in depicting the actual software behaviors

Sequence squeezing was not used

Seungyeon et al. 2021 33

Two stage hybrid malware detection

Provides better accuracy

Takes more time in feature extraction

Chen et al. 202134

Recurrent neural networks with LSTM and GRU

Provides better classification and prediction with time-series

Accuracy was low

Miao et al. 202335

Time –controllable keyword search scheme

Improves efficiency

Issue in security

Miao et al. 202 36

PSDQ

Reduces the query time

Less security