Comparing the Forecast Performance of Advanced Statistical and Machine Learning Techniques Using Huge Big Data: Evidence from Monte Carlo Experiments

Table 4

Methods' performance across various scenarios.

Scenarios	Three cases
	Low	Medium	High
Multicollinearity	E-SCAD is best under a small sample. MCP is the best option in case of a large sample.	E-SCAD is best under a small sample. MCP is the best option in case of a large sample.	PLS-based factor model provides a better forecast under small sample. In case of a large sample, E-SCAD is superior.
Heteroscedasticity	MCP is best.	MCP is best.	MCP is best.
Autocorrelation	E-SCAD is best under a small sample. MCP is the best option using more data.	E-SCAD is best under a small sample. MCP is the best option using more data.	E-SCAD is best under a small sample. MCP is the best option using a large data set.